A method for processing map feature data of roadside equipment

Abstract

This invention relates to a method for processing map element data for roadside equipment. The method includes: the roadside equipment performing static target identification and target attribute identification, and updating the static target tracking sequence set based on the identification results; performing road surface condition target identification, and updating the road surface target tracking sequence set based on the identification results; identifying whether any static road targets have undergone state changes; updating the corresponding map elements in the vector map based on the changed target sequence; publishing local data based on the updated element sequence; synchronizing with the cloud platform based on the updated element sequence and the overall scenic spot cloud; identifying road events based on the vector map and publishing the event list locally; and identifying road surface events based on the road surface target tracking sequence set and the vector map, publishing the event list locally and synchronizing with the cloud platform. This invention can improve the timeliness of roadside map updates and the personalization level of vehicle-road cooperation while reducing the pressure on the cloud platform.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a map element data processing method of a roadside device. BACKGROUND

[0002] In the application scenarios of Internet of Vehicles and Car-Road Cooperation, the roadside devices (Road Side Unit, RSU) placed at the positions of roadsides and road intersections are mainly responsible for receiving the sensing data sent by the front-end sensing devices (cameras, radars) and forwarding the sensing data to the remote cloud platform, and receiving the roadside map information pushed by the remote cloud platform and pushing the roadside map information to the passing vehicles. The roadside devices in the conventional scheme are only used for data forwarding and do not participate in the data processing of the roadside map. With the in-depth development of the application scenarios of Internet of Vehicles and Car-Road Cooperation, it is found that the conventional scheme is affected by the cloud platform resources, the real-time communication efficiency of the cloud platform and other factors, and it is difficult to ensure that all roadside devices can timely receive the latest roadside map information. SUMMARY

[0003] The purpose of the present application is to provide a map element data processing method of a roadside device, an electronic device and a computer readable storage medium, which overcomes the defects of the prior art. The roadside device tracks the local road static targets (intersection, road, road boundary, lane, lane marking, pole-shaped object, traffic sign, arrow, U-turn area, stop line, signal light, pedestrian crossing, no-entry area in intersection, safety island, left-turn waiting area, flow guide area, obstacle, etc.) and road surface conditions (accident, spill, accumulated water, accumulated snow, icing, crack, pothole, collapse, construction, etc.) according to the visual sensing data, confirms whether the map elements change based on the static target tracking information, and performs real-time map updating and message publishing locally when the change is confirmed, and synchronizes the updated information to the cloud platform. The roadside device identifies the road surface events (road surface accident, road surface spill, road surface accumulated water, road surface accumulated snow, road surface icing, road surface crack, road surface pothole, road surface collapse, road surface construction, etc.) in the current area based on the road surface condition tracking information, combines the identification result with the map elements, and outputs the corresponding event list for real-time message publishing and remote cloud platform synchronization. The roadside device identifies the road events (timed speed limit event, timed lane restriction event, tidal lane event, etc.) according to the local map, combines the identification result with the map elements, and outputs the corresponding event list for real-time message publishing. The roadside device of the present application can update the local map in real time without relying on the cloud platform, can synchronize the latest updated data to the cloud platform, and can provide more rich local traffic information to the local traffic participants. Through the present application, the timeliness of roadside map updating can be improved, the individualization level of Car-Road Cooperation can be improved, and the data processing pressure of the cloud platform can be reduced.

[0004] To achieve the above object, the embodiment of the present application provides a kind of map feature data processing method of roadside equipment, the method comprises:

[0005] The first roadside equipment receives the monitoring range vector map of roadside equipment issued by the first cloud platform in advance as the corresponding first vector map and saves;And at any time, the first image issued by all visual perception devices forms the corresponding first image sequence, and the first point cloud issued by all radar perception devices forms the corresponding first point cloud sequence;And according to the latest first image sequence, the roadside panoramic map splicing is generated corresponding to the first panoramic map;And according to the latest first point cloud sequence, the roadside panoramic point cloud fusion is generated corresponding to the first panoramic point cloud;

[0006] According to the first panoramic map, road static target identification and target attribute identification are carried out, and the static target tracking sequence set is updated according to the identification result;And according to the first panoramic map, road surface condition target identification is carried out, and the road surface target tracking sequence set is updated according to the identification result;

[0007] According to the static target tracking sequence set, whether all road static targets change state is identified to generate the corresponding first change target sequence;And according to the first change target sequence, the corresponding map feature in the first vector map is updated to generate the corresponding first update element sequence;And according to the first update element sequence, local data is published;And according to the first update element sequence and the first panoramic point cloud, cloud platform data synchronization is carried out;

[0008] According to the first vector map, road event identification is carried out to generate the corresponding first event list, and local data is published according to the first event list;And according to the road surface target tracking sequence set and the first vector map, road surface event identification is carried out to generate the corresponding second event list, and local data is published and cloud platform data synchronization is carried out according to the second event list.

[0009] Preferably, the first vector map is the vector map of the roadside area corresponding to the first roadside equipment;The first vector map includes a first map version and a plurality of first map features;The first map feature includes a first feature number, a first feature type, a first feature coordinate, a first feature geometric parameter and a first feature attribute set;The first feature type includes intersection, road, road boundary, lane, lane marking, pole, traffic sign, arrow, U-turn area, stop line, signal light, pedestrian crossing, intersection forbidden area, safety island, left turn waiting area, flow guide area and obstacle;

[0010] The device type of the visual perception device includes gun camera and fish eye camera;

[0011] The device type of the radar perception device includes millimeter wave radar and laser radar;

[0012] The static target tracking sequence set includes a plurality of first target tracking sequences; the first target tracking sequence includes a first target number, a first target type, and a plurality of first target tracking data; the first target type includes an intersection, a road, a road boundary, a lane, a lane marking, a pole, a traffic sign, an arrow, a U-turn area, a stop line, a signal light, a pedestrian crossing, a no-entry area in an intersection, a safety island, a left-turn waiting area, a guide flow area, and an obstacle; the first target tracking data includes first target coordinates, first target geometric parameters, a first target attribute set, and a first timestamp;

[0013] The road surface target tracking sequence set includes a plurality of second target tracking sequences; the second target tracking sequence includes a second target number, a second target type, and a plurality of second target tracking data; the second target type includes an accident road surface, a spilled road surface, a water accumulation road surface, a snow accumulation road surface, an icy road surface, a crack road surface, a pothole road surface, a collapsed road surface, and a construction road surface; the second target tracking data includes second target coordinates, a first target range vertex coordinate set, and a second timestamp;

[0014] The first change target sequence includes one or more first change targets; the first change target includes a third target number, a third target type, a first change type, a first change before coordinate, a first change after coordinate, a first change before geometric parameter, a first change after geometric parameter, a first change before attribute set, a first change after attribute set, and a third timestamp; the first change type includes an addition type, a deletion type, a movement type, and an in-place change type;

[0015] The first update element sequence includes a plurality of first update elements; the first update element includes a second element number, a second element type, a first element update type, a first element update before coordinate, a first element update after coordinate, a first element update before geometric parameter, a first element update after geometric parameter, a first element update before attribute set, a first element update after attribute set, and a first element update timestamp;

[0016] The first event list includes a plurality of first event records; the first event record includes a first event name, a first event lane number, a first event valid time period and a first event parameter; the first event name includes a speed limit event, a traffic limit event and a tidal lane event; when the first event name is a speed limit event, the corresponding first event valid time period is a speed limit time period, and the corresponding first event parameter is a speed limit threshold; when the first event name is a traffic limit event, the corresponding first event valid time period is a traffic limit time period, and the corresponding first event parameter is a traffic limit vehicle type set; when the first event name is a tidal lane event, the corresponding first event valid time period is a lane changing time period, and the corresponding first event parameter is a road driving direction after lane changing;

[0017] The second event list includes a plurality of second event records; the second event record includes a second event name, a first event type, a first event center coordinate, a first event range vertex coordinate set, a first event range lane number and a first event range lane number set; the second event name includes a road accident event, a road spill event, a road water accumulation event, a road snow accumulation event, a road icing event, a road crack event, a road pit event, a road collapse event and a road construction event; the first event type includes an added type, a range reduced type, a range expanded type and an end type; when the first event type is the added type, the range reduced type or the range expanded type, a set of information composed of the corresponding first event center coordinate, the first event range vertex coordinate set, the first event range lane number and the first event range lane number set is the latest set of real-time information corresponding to the current event; when the first event type is the end type, a set of information composed of the corresponding first event center coordinate, the first event range vertex coordinate set, the first event range lane number and the first event range lane number set is the last set of historical information corresponding to the current event.

[0018] Preferably, the road static target recognition and target attribute recognition are performed according to the first panoramic map, and a static target tracking sequence set is updated according to a recognition result, and the static target recognition and target attribute recognition specifically include:

[0019] The first panoramic image is subjected to road static target recognition and classification processing to obtain a plurality of corresponding first target recognition boxes; the image area covered by each first target recognition box on the first panoramic image is extracted as a corresponding first target image; and the time information corresponding to the first panoramic image is taken as a corresponding first time stamp; the first target recognition box includes a first recognition box center point coordinate, a first recognition box geometric size, and a first recognition box target type; the first recognition box target type includes an intersection, a road, a road boundary, a lane, a lane marking, a rod-shaped object, a traffic sign, an arrow, a U-turn area, a stop line, a signal light, a pedestrian crossing, an in-intersection no-entry area, a safety island, a left-turn waiting area, a flow guide area, and an obstacle;

[0020] Based on the preset panoramic camera internal and external parameters, the first recognition box center point coordinate of each first target recognition box is subjected to a coordinate information conversion from a pixel coordinate system to a world coordinate system; and based on the preset panoramic camera internal and external parameters, the first recognition box geometric size of each first target recognition box is subjected to a size information conversion from an image size to an actual size;

[0021] traverse each of the first target recognition boxes; and in the traversal, take the first target recognition box currently traversed as a corresponding current recognition box; and take the first target tracking sequence in the static target tracking sequence set that matches the first target type of the current recognition box as a corresponding first preselected sequence; and calculate the straight-line distance between the first target coordinates of the last first target tracking data in each of the first preselected sequences and the first recognition box center point coordinates of the current recognition box to obtain a corresponding first straight-line distance, and take the shortest distance among them as a corresponding shortest straight-line distance; and identify whether the shortest straight-line distance exceeds a preset first shortest distance threshold; if not, take the first preselected sequence corresponding to the shortest straight-line distance as a corresponding first matching sequence; if so, add a first target tracking sequence in the static target tracking sequence set as a corresponding first matching sequence, assign a unique number to the first matching sequence as the first target number of the first matching sequence, and take the first recognition box target type of the current recognition box as the first target type of the first matching sequence; and after obtaining the first matching sequence, take the first recognition box center point coordinates, the first recognition box geometric size, and the first timestamp of the current recognition box as the first target coordinates, the first target geometric parameters, and the first timestamp; and perform static target attribute extraction processing on the first target image corresponding to the current recognition box to obtain a corresponding first target attribute set; and add a corresponding first target tracking data composed of the first target coordinates, the first target geometric parameters, the first target attribute set, and the first timestamp to the first matching sequence;

[0022] confirm whether the first target tracking sequence is newly added in the current processing process; if yes, each first target tracking sequence newly added in the current processing process is taken as a corresponding first newly added sequence; all the first newly added sequences are traversed; in the traversal, the first newly added sequence currently traversed is taken as a corresponding current newly added sequence, and the first target type of the current newly added sequence is taken as a corresponding current target type; when the current target type is one of a plurality of preset movable target types, other first target tracking sequences in the static target tracking sequence set that match the first target type and the current target type and are not the current newly added sequence are recorded as corresponding preselected same type sequences; when the number of the preselected same type sequences is not 0, the last first target tracking data of each preselected same type sequence is recorded as corresponding preselected target data, and the first target tracking data of the current newly added sequence is taken as corresponding current target data; the first geometric term difference proportion set is obtained by performing difference proportion calculation on each geometric term value in the first target geometric parameter of each preselected target data and the current target data; the first attribute term difference value set is obtained by performing attribute numerical conversion on each attribute term in the first target attribute set of each preselected target data and the current target data and performing difference calculation based on the numerical attribute; the preselected target data for which each first geometric term difference proportion in the first geometric term difference proportion set does not exceed a preset difference proportion threshold and each first attribute term difference value in the first attribute term difference value set is 0 is recorded as corresponding matching target data; if the number of the matching target data is 1, the current target data is added to the first target tracking sequence corresponding to the matching target data, and the current newly added sequence is deleted from the static target tracking sequence set; if the number of the matching target data is greater than 1, the first geometric term difference proportion set of each matching target data is taken as a corresponding first geometric term difference proportion set, the first mean proportion is obtained by performing mean value calculation on all the first geometric term difference proportions in the first geometric term difference proportion set, the matching target data corresponding to the smallest first mean proportion is taken as corresponding final matching target data, the current target data is added to the first target tracking sequence corresponding to the final matching target data, and the current newly added sequence is deleted from the static target tracking sequence set; wherein the plurality of movable target types include a rod-shaped object, a traffic sign, a signal lamp, a safety island and an obstacle; the first geometric term difference proportion set includes a plurality of first geometric term difference proportions, and a first geometric term difference proportion = |geometric term value of current target data - geometric term value of preselected target data| / geometric term value of preselected target data;The first attribute item difference score value set includes a plurality of first attribute item difference scores, and the first attribute item difference score=(attribute item value of current target data-preselected attribute item value of target data).

[0023] Preferably, the road surface condition target identification according to the first panoramic image and updating the road surface target tracking sequence set according to the identification result specifically includes:

[0024] The road surface condition target identification and classification processing of the first panoramic image obtain a plurality of corresponding second target identification boxes; and the time information corresponding to the first panoramic image is taken as the corresponding second time stamp; the second target identification box includes a second identification box center point coordinate, a second identification box geometric size, and a second identification box target type; the second identification box target type includes an accident road surface, a spill road surface, a water accumulation road surface, a snow accumulation road surface, an icy road surface, a crack road surface, a pothole road surface, a collapse road surface, and a construction road surface;

[0025] The driving road surface area identification of the first panoramic image obtains a corresponding first road surface area; and the intersection surface of each second target identification box with the first road surface area is taken as a corresponding first intersection surface; the convex polygon conversion of each first intersection surface obtains a corresponding first convex polygon; and all vertex coordinates of each first convex polygon are extracted to form a corresponding first convex polygon vertex coordinate set;

[0026] Based on the preset panoramic image camera internal and external parameters, the second identification box center point coordinate of each second target identification box is converted from a pixel coordinate system to a world coordinate system; and based on the preset panoramic image camera internal and external parameters, each vertex coordinate of the first convex polygon vertex coordinate set corresponding to each second target identification box is converted from a pixel coordinate system to a world coordinate system;

[0027] traverse each of the second target recognition boxes; and in the traversal, take the second target recognition box currently traversed as a corresponding current recognition box; and take the second target tracking sequence in the road surface target tracking sequence set that matches the second target type of the second target tracking sequence with the second recognition box target type of the current recognition box as a corresponding second preselected sequence; and when the number of second preselected sequences is 0, add a second target tracking sequence in the road surface target tracking sequence set as a corresponding second matching sequence, assign a unique number to the second matching sequence as the second target number of the second matching sequence, and take the second recognition box target type of the current recognition box as the second target type of the second matching sequence; and when the number of second preselected sequences is greater than 0, calculate the straight-line distance between the second target coordinates of the last second target tracking data in each of the second preselected sequences and the second recognition box center point coordinates of the current recognition box to obtain a corresponding second straight-line distance, take the shortest of them as the shortest straight-line distance, and identify whether the shortest straight-line distance exceeds a preset second shortest distance threshold, if not, take the second preselected sequence corresponding to the shortest straight-line distance as the second matching sequence, if so, add a second target tracking sequence in the road surface target tracking sequence set as a corresponding second matching sequence, assign a unique number to the second matching sequence as the second target number of the second matching sequence, and take the second recognition box target type of the current recognition box as the second target type of the second matching sequence; and after obtaining the second matching sequence, take the second recognition box center point coordinates of the current recognition box, the first convex polygon vertex coordinate set, and the second timestamp as the second target coordinates, the first target range vertex coordinate set, and the second timestamp to form a corresponding second target tracking data to the second matching sequence.

[0028] Preferably, the identification of whether all road static targets have state changes according to the static target tracking sequence set generates a corresponding first change target sequence, specifically including:

[0029] The number of first target tracking data of each first target tracking sequence in the static target tracking sequence set is counted to generate a corresponding first number; and the first target tracking sequence with the first number of 1 is recorded as a corresponding new sequence; and the first target tracking sequence with the first number greater than 1 is recorded as a corresponding inventory sequence.

[0030] and the first time stamp of the last first target tracking data of each of the inventory sequences is identified whether it is earlier than the current time stamp; if yes, the current inventory sequence is recorded as a corresponding stationary sequence; if no, the current inventory sequence is recorded as a corresponding continuous sequence;

[0031] and the first target coordinates of the last two first target tracking data of each of the continuous sequences are extracted as corresponding first and second coordinates; and whether the straight line distance between the first and second coordinates exceeds a preset first minimum distance threshold is identified; if yes, the current continuous sequence is recorded as a corresponding moving sequence;

[0032] and the first time stamp of the second last first target tracking data of any of the remaining continuous sequences is extracted as a corresponding reference time stamp; and whether the first time stamp of the last first target tracking data of each of the stationary sequences is earlier than the reference time stamp is identified; if yes, the current stationary sequence is recorded as a corresponding long-term stationary sequence; if no, the current stationary sequence is recorded as a corresponding current stationary sequence;

[0033] and the first change target sequence is initialized as an empty sequence;

[0034] and when the number of the added sequences is not 0, traversing each of the added sequences; and when traversing, taking the currently traversed added sequence as a corresponding current sequence; and setting a corresponding third target number as the first target number of the current sequence, setting a corresponding third target type as the first target type of the current sequence, setting a corresponding first change type as an added type, setting a corresponding first change before coordinate as empty, setting a corresponding first change after coordinate as the first target coordinate of the first first target tracking data of the current sequence, setting a corresponding first change before geometric parameter as empty, setting a corresponding first change after geometric parameter as the first target geometric parameter of the first first target tracking data of the current sequence, setting a corresponding first change before attribute set as empty, setting a corresponding first change after attribute set as the first target attribute set of the first first target tracking data of the current sequence, setting a corresponding third timestamp as the first timestamp of the first first target tracking data of the current sequence; and adding a corresponding first change target to the first change target sequence from the obtained third target number, third target type, first change type, first change before coordinate, first change after coordinate, first change before geometric parameter, first change after geometric parameter, first change before attribute set, first change after attribute set, and third timestamp;

[0035] and when the number of the current period stationary sequences is not 0, traversing each of the current period stationary sequences; and when traversing, taking the current period stationary sequence being traversed as a corresponding current sequence; and setting a corresponding third target number as the first target number of the current sequence, setting a corresponding third target type as the first target type of the current sequence, setting a corresponding first change type as a deletion type, setting a corresponding first change before coordinate as the first target coordinate of the last first target tracking data of the current sequence, setting a corresponding first change after coordinate as null, setting a corresponding first change before geometric parameter as the first target geometric parameter of the last first target tracking data of the current sequence, setting a corresponding first change after geometric parameter as null, setting a corresponding first change before attribute set as the first target attribute set of the last first target tracking data of the current sequence, setting a corresponding first change after attribute set as null, and setting a corresponding third timestamp as time information corresponding to the first panoramic picture; and adding a corresponding first change target into the first change target sequence by the third target number, the third target type, the first change type, the first change before coordinate, the first change after coordinate, the first change before geometric parameter, the first change after geometric parameter, the first change before attribute set, the first change after attribute set, and the third timestamp;

[0036] and when the number of the movement sequences is not 0, traversing each of the movement sequences; and when traversing, taking the movement sequence being currently traversed as a corresponding current sequence, and taking the last two first target tracking data of the current sequence as corresponding pre-change tracking data and post-change tracking data; and setting a corresponding third target number as the first target number of the current sequence, setting a corresponding third target type as the first target type of the current sequence, setting a corresponding first change type as a movement type, setting a corresponding first pre-change coordinate as the first target coordinate of the pre-change tracking data, setting a corresponding first post-change coordinate as the first target coordinate of the post-change tracking data, setting a corresponding first pre-change geometric parameter as the first target geometric parameter of the pre-change tracking data, setting a corresponding first post-change geometric parameter as the first target geometric parameter of the post-change tracking data, setting a corresponding first pre-change attribute set as the first target attribute set of the pre-change tracking data, setting a corresponding first post-change attribute set as the first target attribute set of the post-change tracking data, and setting a corresponding third timestamp as the first timestamp of the post-change tracking data; and adding, to the first change target sequence, a corresponding first change target composed of the third target number, the third target type, the first change type, the first pre-change coordinate, the first post-change coordinate, the first pre-change geometric parameter, the first post-change geometric parameter, the first pre-change attribute set, the first post-change attribute set, and the third timestamp;

[0037] and each of the first target geometric parameters of the first and second tracking data is subjected to difference ratio calculation to obtain a corresponding second geometric term difference ratio set, and each of the first target attribute sets of the first and second tracking data is subjected to attribute numerical conversion and difference calculation based on the numerical attribute to obtain a corresponding second attribute term difference value set; and whether all second geometric term difference ratios of the second geometric term difference ratio set are less than a preset difference ratio threshold is identified, and if yes, a corresponding first identification result is set as no change in geometric parameters, and if not, the corresponding first identification result is set as change in geometric parameters; and whether all first attribute term difference values of the second attribute term difference value set are 0 is identified, and if yes, a corresponding second identification result is set as no change in attribute, and if not, the corresponding second identification result is set as change in attribute; and when the first identification result is change in geometric parameters or the second identification result is change in attribute, a corresponding third target number is set as the first target number of the current sequence, a corresponding third target type is set as the first target type of the current sequence, a corresponding first change type is set as a stationary change type, a corresponding first change before coordinate is set as the first target coordinate of the first tracking data, a corresponding first change after coordinate is set as the first target coordinate of the second tracking data, a corresponding first change before geometric parameter is set as the first target geometric parameter of the first tracking data, a corresponding first change after geometric parameter is set as the first target geometric parameter of the second tracking data, a corresponding first change before attribute set is set as the first target attribute set of the first tracking data, a corresponding first change after attribute set is set as the first target attribute set of the second tracking data, and a corresponding third timestamp is set as the first timestamp of the second tracking data, and a corresponding first change target is added to the first change target sequence by the third target number, the third target type, the first change type, the first change before coordinate, the first change after coordinate, the first change before geometric parameter, the first change after geometric parameter, the first change before attribute set, the first change after attribute set, and the third timestamp; wherein the second geometric term difference ratio set includes a plurality of second geometric term difference ratios, and a second geometric term difference ratio = |geometric term value of second tracking data - geometric term value of first tracking data| / geometric term value of first tracking data.The second attribute item differential value set includes a plurality of second attribute item differential values, and the second attribute item differential value=(attribute item value of the second tracking data-first attribute item value of the first tracking data).

[0038] Preferably, the updating of the corresponding map elements in the first vector map according to the first change target sequence to generate a corresponding first update element sequence specifically includes:

[0039] when the first change target sequence is not empty, initializing the first update element sequence as an empty sequence; and traversing each first change target of the first change target sequence; and when traversing, taking the first change target being currently traversed as a corresponding current change target, taking the third target type of the current change target as a corresponding current target type, taking the first change before coordinate, the first change before geometric parameter and the first change before attribute set of the current change target as a corresponding before coordinate, before geometric parameter and before attribute set; and taking, through querying the first vector map, the first map element whose first element type matches the current target type and whose first element coordinate, first element geometric parameter and first element attribute set respectively match the corresponding before coordinate, before geometric parameter and before attribute set as a corresponding current map element; and when the current map element is not empty, setting the corresponding second element number as the first element number of the current map element, setting the corresponding second element type as the first element type of the current map element, setting the corresponding first element update type as the first change type of the current change target, setting the corresponding first element update before coordinate as the first change before coordinate of the current change target, setting the corresponding first element update after coordinate as the first change after coordinate of the current change target, setting the corresponding first element update before geometric parameter as the first change before geometric parameter of the current change target, setting the corresponding first element update after geometric parameter as the first change after geometric parameter of the current change target, setting the corresponding first element update before attribute set as the first change before attribute set of the current change target, setting the corresponding first element update after attribute set as the first change after attribute set of the current change target, setting the corresponding first element update timestamp as the third timestamp of the current change target; and adding a corresponding first update element composed of the obtained second element number, second element type, first element update type, first element update before coordinate, first element update after coordinate, first element update before geometric parameter, first element update after geometric parameter, first element update before attribute set, first element update after attribute set and first element update timestamp into the first update element sequence; and when the sequence element is added successfully, updating the first element coordinate, first element geometric parameter and first element attribute set corresponding to the current map element according to the first change after coordinate, first change after geometric parameter and first change after attribute set of the current change target.

[0040] Preferably, the local data publishing according to the first update element sequence specifically comprises:

[0041] publishing the first update element sequence into a locally preset roadside map element update message queue for message subscription by any traffic participant client connected with the current first roadside device; the traffic participant client includes a vehicle-mounted driving system of any type of motor vehicle, a travel navigation device, equipment or system of any type of non-motor vehicle, and a navigation client, device, equipment or system of any type of personal user.

[0042] Preferably, the cloud platform data synchronization according to the first update element sequence and the first panoramic point cloud specifically includes:

[0043] when the first update element sequence is not empty, recording the first update element of the second element type as one of the preset multiple types of static core element types as a corresponding core element, and recording the first update element of the second element type not matching any type of the preset multiple types of static core element types as a corresponding non-core element in the first update element sequence;

[0044] and all the non-core elements form a corresponding first update element sequence;

[0045] and when the first panoramic point cloud is not empty, performing point cloud target recognition on the first panoramic point cloud to obtain a plurality of third target recognition boxes; and taking the point cloud in each of the third target recognition boxes of the first panoramic point cloud as a corresponding first target point cloud; the third target recognition box includes a third recognition box center point coordinate and a third recognition box geometric size;

[0046] and iterating each of the core elements; and when iterating, taking the currently iterated core element as a corresponding current element, and taking the first changed coordinate of the current element as a corresponding current element coordinate; and calculating the straight line distance between the current element coordinate and each of the third recognition box center point coordinates to obtain a corresponding second straight line distance; and selecting the minimum value from all the obtained second straight line distances as a corresponding third straight line distance; and identifying whether the third straight line distance exceeds a preset third minimum distance threshold, if not, taking the first target point cloud corresponding to the third straight line distance as a corresponding first element updated point cloud, if so, setting the corresponding first element updated point cloud as empty; and taking the current element and the corresponding first element updated point cloud to form a corresponding second update element; and when the iteration ends, taking all the obtained second update elements to form a corresponding second update element sequence;

[0047] extracting the locally preset roadside device number as a corresponding first roadside device number; extracting the locally preset intersection or road number corresponding to the first roadside device number as a corresponding first intersection / road number; extracting the time information of the first panoramic map as a corresponding first report timestamp; and composing a corresponding first perception device type set by the device types of all the visual perception devices corresponding to the first panoramic map; composing a corresponding first perception device number set by the device numbers of all the visual perception devices corresponding to the first panoramic map; composing a corresponding second perception device type set by the device types of all the radar perception devices corresponding to the first panoramic point cloud; composing a corresponding second perception device number set by the device numbers of all the radar perception devices corresponding to the first panoramic point cloud; and extracting the first target tracking sequence corresponding to each first update element in the first update element sequence in the static target tracking sequence set to compose a corresponding first target tracking sequence set;

[0048] and sending the corresponding first cloud platform synchronization data report to the first cloud platform by the obtained first roadside device number, first intersection / road number, first map version of the first vector map, first panoramic map, first report timestamp, first perception device type set, first perception device number set, second perception device type set, second perception device number set, first target tracking sequence set, first update element sequence and second update element sequence.

[0049] Preferably, the road event identification according to the first vector map generates a corresponding first event list, specifically including:

[0050] The first map element of the first element type of lane in the first vector map is recorded as a corresponding first lane element; and the first event list is initialized as empty;

[0051] and whether a speed limit attribute exists in the first element attribute set of each first lane element; if so, a corresponding first speed limit time period and a first speed limit threshold are extracted from the speed limit attribute; and whether the current time is not earlier than the start time of the first speed limit time period is identified; if so, the corresponding first event name is set as a speed limit event, the corresponding first event lane number is set as the first element number of the current first lane element, the corresponding first event valid time period is set as the corresponding first speed limit time period, the corresponding first event parameter is set as the first speed limit threshold, and a corresponding first event record composed of the obtained first event name, first event lane number, first event valid time period and first event parameter is added to the first event list;

[0052] and whether a speed limit attribute exists in the first element attribute set of each first lane element; if so, a corresponding first speed limit time period and a first speed limit threshold are extracted from the speed limit attribute; and whether the current time is not earlier than the start time of the first speed limit time period is identified; if so, the corresponding first event name is set as a speed limit event, the corresponding first event lane number is set as the first element number of the current first lane element, the corresponding first event valid time period is set as the corresponding first speed limit time period, the corresponding first event parameter is set as the first speed limit threshold, and a corresponding first event record composed of the obtained first event name, first event lane number, first event valid time period and first event parameter is added to the first event list;

[0053] and whether a speed limit attribute exists in the first element attribute set of each first lane element; if so, a corresponding first speed limit time period and a first speed limit threshold are extracted from the speed limit attribute; and whether the current time is not earlier than the start time of the first speed limit time period is identified; if so, the corresponding first event name is set as a speed limit event, the corresponding first event lane number is set as the first element number of the current first lane element, the corresponding first event valid time period is set as the corresponding first speed limit time period, the corresponding first event parameter is set as the first speed limit threshold, and a corresponding first event record composed of the obtained first event name, first event lane number, first event valid time period and first event parameter is added to the first event list.

[0054] Preferably, the local data publishing according to the first event list specifically includes:

[0055] publish the first event list to a locally preset lane time-limited event message queue for message subscription by any traffic participant client connected with the current first roadside device; the traffic participant client includes a vehicle-mounted driving system of any type of motor vehicle, a travel navigation device, equipment or system of any type of non-motor vehicle, and a navigation client, device, equipment or system of any type of personal user.

[0056] Preferably, the road event identification according to the road target tracking sequence set and the first vector map generates a corresponding second event list, specifically including:

[0057] When the road target tracking sequence set is not empty, the number of the second target tracking data of each second target tracking sequence of the road target tracking sequence set is counted to generate a corresponding third number; and the second target tracking sequence with the third number of 1 is recorded as a corresponding new sequence; and the second target tracking sequence with the third number greater than 1 is recorded as a corresponding inventory sequence;

[0058] The time information corresponding to the first panoramic map is taken as a current timestamp; and whether the second timestamp of the last second target tracking data of each inventory sequence is earlier than the current timestamp is identified; if yes, the current inventory sequence is recorded as a corresponding stationary sequence; if not, the current inventory sequence is recorded as a corresponding continuous sequence;

[0059] The first and second vertex coordinate sets of the first target range vertex coordinate set of the last two first target tracking data of each continuous sequence are extracted as corresponding first and second vertex coordinate sets; and the first and second estimated areas are obtained by estimating the area of the road convex polygon according to the first and second vertex coordinate sets; and the first area difference and the first area change ratio are calculated by calculating the area change difference and the area change ratio according to the first and second estimated areas, first area difference = second estimated area - first estimated area, first area change ratio = first area difference / first estimated area; and whether the absolute value of the first area change ratio exceeds a preset change ratio threshold is identified; if not, the current continuous sequence is recorded as a corresponding range unchanged sequence; if yes, the current continuous sequence is recorded as a corresponding range larger sequence when the first area difference is greater than 0, and the current continuous sequence is recorded as a corresponding range smaller sequence when the first area difference is less than 0;

[0060] The second event list is initialized as an empty sequence;

[0061] and when the number of the added sequences is not 0, traversing each of the added sequences; and when traversing, taking the currently traversed added sequence as a corresponding current sequence, taking the first second target tracking data of the current sequence as a corresponding current target tracking data, and taking the second target type of the current sequence as a corresponding current target type; and when the current target type is an accident road surface, a spilled road surface, a water accumulated road surface, a snow accumulated road surface, an icy road surface, a crack road surface, a pothole road surface, a collapsed road surface, or a construction road surface, setting a corresponding second event name as a corresponding road accident event, a road spilled event, a water accumulated road event, a snow accumulated road event, an icy road event, a crack road event, a pothole road event, a collapsed road event, or a construction road event; setting a corresponding first event type as an added type; setting a corresponding first event center coordinate as the second target coordinate of the current target tracking data; setting a corresponding first event range vertex coordinate set as the first target range vertex coordinate set of the current target tracking data; recording, as a corresponding first event lane element, a first map element in the first vector map, where a first element type is a lane and a lane address range attribute in a first element attribute set corresponds to a road area intersected with a road convex polygon area corresponding to the first event range vertex coordinate set; counting a number of the first event lane elements to generate a corresponding first lane number, and setting a corresponding first event range lane number as the corresponding first lane number; and setting a corresponding first event range lane number set as a first event lane number set composed of the first element numbers of all the first event lane elements; and adding, to the second event list, a corresponding second event record composed of the obtained second event name, first event type, first event center coordinate, first event range vertex coordinate set, first event range lane number, and first event range lane number set;

[0062] and when the number of the range-reducing sequences is not 0, traversing each of the range-reducing sequences; and when traversing, taking the currently traversed range-reducing sequence as a corresponding current sequence, taking the last second target tracking data of the current sequence as a corresponding current target tracking data, and taking the second target type of the current sequence as a corresponding current target type; and when the current target type is an accident road surface, a spilled road surface, a water-accumulated road surface, a snow-accumulated road surface, an icy road surface, a cracked road surface, a pothole road surface, a collapsed road surface, or a construction road surface, setting a corresponding first event name as a corresponding road accident event, a road spilled event, a water-accumulated road event, a snow-accumulated road event, an icy road event, a cracked road event, a pothole road event, a collapsed road event, or a construction road event; setting a corresponding first event type as a range-reducing type; setting a corresponding first event center coordinate as the second target coordinate of the current target tracking data; setting a corresponding first event range vertex coordinate set as the first target range vertex coordinate set of the current target tracking data; and recording, as a corresponding second event lane element, a first map element corresponding to a road surface area in the first vector map, where the first element type is a lane and a lane address range attribute in the first element attribute set, and a road convex polygon area corresponding to the first event range vertex coordinate set; counting the number of the first event lane elements to generate a corresponding second lane number, and setting a corresponding first event range lane number as the second lane number; and setting a corresponding first event range lane number set as a corresponding second event lane number set composed of the first element numbers of all the second event lane elements; and adding, to the second event list, a corresponding second event record composed of the obtained second event name, first event type, first event center coordinate, first event range vertex coordinate set, first event range lane number, and first event range lane number set.

[0063] and when the number of the range-increasing sequences is not 0, traversing each of the range-increasing sequences; and when traversing, taking the currently traversed range-increasing sequence as a corresponding current sequence, taking the last second target tracking data of the current sequence as a corresponding current target tracking data, and taking the second target type of the current sequence as a corresponding current target type; and when the current target type is an accident road surface, a spilled road surface, a water-accumulated road surface, a snow-accumulated road surface, an icy road surface, a cracked road surface, a pothole road surface, a collapsed road surface, or a construction road surface, setting a corresponding second event name as a corresponding road surface accident event, a road surface spilled event, a water-accumulated road surface event, a snow-accumulated road surface event, an icy road surface event, a cracked road surface event, a pothole road surface event, a collapsed road surface event, or a construction road surface event; setting a corresponding first event type as a range expansion type; setting a corresponding first event center coordinate as the second target coordinate of the current target tracking data; setting a corresponding first event range vertex coordinate set as the first target range vertex coordinate set of the current target tracking data; and recording, as a corresponding third event lane element, a first map element corresponding to a road surface area in the first vector map, where the first element type is a lane and a lane address range property in the first element attribute set, and a road surface convex polygon area corresponding to the first event range vertex coordinate set; counting the number of the third event lane elements to generate a corresponding third lane number, and setting a corresponding first event range lane number as the third lane number; and setting a corresponding first event range lane number set as a third event lane number set composed of the first element numbers of all the third event lane elements; and adding, to the second event list, a corresponding second event record composed of the obtained second event name, first event type, first event center coordinate, first event range vertex coordinate set, first event range lane number, and first event range lane number set.

[0064] and when the number of the stagnation sequences is not 0, traversing each of the stagnation sequences; and when traversing, taking the stagnation sequence currently traversed as a corresponding current sequence, taking the last second target tracking data of the current sequence as a corresponding current target tracking data, and taking the second target type of the current sequence as a corresponding current target type; and when the current target type is an accident road surface, a spilled road surface, a water accumulated road surface, a snow accumulated road surface, an icy road surface, a crack road surface, a pothole road surface, a collapsed road surface, or a construction road surface, setting a corresponding second event name as a corresponding road surface accident event, a road surface spilled event, a water accumulated road surface event, a snow accumulated road surface event, an icy road surface event, a crack road surface event, a pothole road surface event, a collapsed road surface event, or a construction road surface event; setting a corresponding first event type as an end type; setting a corresponding first event center coordinate as the second target coordinate of the current target tracking data; setting a corresponding first event range vertex coordinate set as the first target range vertex coordinate set of the current target tracking data; recording, as a corresponding fourth event lane element, a first map element corresponding to a road surface area of a lane address range attribute in the first element attribute set of the first element type of lane in the first vector map and a road surface convex polygon area corresponding to the first event range vertex coordinate set; counting the number of the fourth event lane elements to generate a corresponding fourth lane number, and setting a corresponding first event range lane number as the fourth lane number; composing a corresponding fourth event lane number set from the first element numbers of all the fourth event lane elements, and setting a corresponding first event range lane number set as the fourth event lane number set; and adding a corresponding second event record composed of the second event name, the first event type, the first event center coordinate, the first event range vertex coordinate set, the first event range lane number, and the first event range lane number set to the second event list; and deleting the current sequence from the road surface target tracking sequence set when the list record is added successfully.

[0065] Preferably, the local data publishing and cloud platform data synchronization according to the second event list specifically includes:

[0066] when the second event list is not empty, publishing the second event record with the second event name as a road surface accident event in the second event list to a locally preset traffic accident event message queue for message subscription by any traffic participant client connected to the current first road side device, and publishing the second event record with the second event name as a road surface construction event in the second event list to a locally preset road construction event message queue for message subscription by any traffic participant client connected to the current first road side device, and publishing the second event record with the second event name as a road surface spillage event, a road surface water event, a road surface snow event and a road surface icing event in the second event list to a locally preset road safety event message queue for message subscription by any traffic participant client connected to the current first road side device, and publishing the second event record with the second event name as a road surface crack event, a road surface pothole event and a road surface collapse event in the second event list to a locally preset road disease event message queue for message subscription by any traffic participant client connected to the current first road side device; the traffic participant client includes a vehicle-mounted driving system of any type of motor vehicle, a travel navigation device, equipment or system of any type of non-motor vehicle, and a navigation client, device, equipment or system of any type of personal user;

[0067] extracting a locally preset road side device number as a corresponding second road side device number, and extracting a locally preset intersection or road number corresponding to the second road side device number as a corresponding second intersection / road number, and taking the time information of the first panoramic map as a corresponding second report timestamp, and sending a corresponding second cloud platform synchronization data report to the first cloud platform by the second road side device number, the second intersection / road number, the first map version of the first vector map, the first panoramic map, the second report timestamp, the road target tracking sequence set and the second event list.

[0068] The second aspect of the embodiment of the application provides an electronic device, including a memory, a processor and a transceiver.

[0069] The processor is used for coupling with the memory, reading and executing instructions in the memory to realize the method steps in the first aspect;

[0070] The transceiver is coupled with the processor, and the transceiver is controlled by the processor to perform message transceiving.

[0071] A third aspect of the present invention provides a computer-readable storage medium storing computer instructions that, when executed by a computer, cause the computer to perform the instructions described in the first aspect.

[0072] This invention provides a method for processing map element data for roadside equipment, an electronic device, and a computer-readable storage medium. The roadside equipment tracks local static road targets (intersections, roads, road boundaries, lanes, lane markings, poles, traffic signs, arrows, U-turn areas, stop lines, traffic lights, pedestrian crossings, no-entry zones within intersections, safety islands, left-turn waiting areas, traffic guidance areas, obstacles, etc.) and road surface conditions (accidents, spills, water accumulation, snow accumulation, icing, cracks, potholes, collapses, construction, etc.) based on the static target tracking information. It then confirms whether map elements have changed based on the static target tracking information and verifies the changes. The system performs real-time map updates and message dissemination locally, synchronizing the updated information to the cloud platform. Based on road condition tracking information, it identifies road events (accidents, spills, water accumulation, snow accumulation, icing, cracks, potholes, collapses, road construction, etc.) in the current area, combining the identification results with map elements to output a corresponding event list for real-time message dissemination and remote cloud platform synchronization. Simultaneously, the roadside equipment identifies road events (timed speed limit events, timed traffic restriction events, tidal lane events, etc.) based on the local map, combining the identification results with map elements to output a corresponding event list for real-time message dissemination. This invention's roadside equipment can update the local map in real-time without relying on a cloud platform, synchronize the latest updated data to the cloud platform, and provide richer local traffic information to local traffic participants. This invention improves the timeliness of roadside map updates, enhances the personalization of vehicle-road cooperation, and reduces the data processing pressure on the cloud platform. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of a map feature data processing method for roadside equipment provided in Embodiment 1 of the present invention;

[0074] Figure 2 This is a schematic diagram of the structure of an electronic device provided in Embodiment 2 of the present invention. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0076] Embodiment 1 of the present invention provides a method for processing map feature data of roadside equipment, such as... Figure 1 The schematic diagram of a map feature data processing method for roadside equipment provided in Embodiment 1 of the present invention mainly includes the following steps:

[0077] Step 1: The first roadside device receives and saves the vector map of the roadside device monitoring range sent by the first cloud platform as the corresponding first vector map; and at any time, it receives the first images sent by all visual perception devices to form the corresponding first image sequence, and receives the first point clouds sent by all radar perception devices to form the corresponding first point cloud sequence; and performs roadside panoramic image stitching to generate the corresponding first panoramic image based on the latest first image sequence; and performs roadside panoramic point cloud fusion to generate the corresponding first panoramic point cloud based on the latest first point cloud sequence.

[0078] The first vector map is a vector map of the roadside area corresponding to the first roadside equipment; the first vector map includes a first map version and multiple first map elements; the first map elements include first element number, first element type, first element coordinates, first element geometric parameters, and first element attribute set; the first element types include intersections, roads, road boundaries, lanes, lane markings, poles, traffic signs, arrows, U-turn areas, stop lines, traffic lights, pedestrian crossings, no-entry areas within intersections, safety islands, left-turn waiting areas, traffic guidance areas, and obstacles; the equipment types of the visual perception equipment include bullet cameras and fisheye cameras; the equipment types of the radar perception equipment include millimeter-wave radar and lidar.

[0079] Here, the first cloud platform of this embodiment of the invention periodically updates the local maps of each first roadside device; the first roadside device, upon receiving the map, stores it locally as the latest base map and updates it accordingly. The first roadside device of this embodiment of the invention also connects to one or more surrounding visual sensing devices / radar sensing devices and receives periodically uploaded sensing data from each device, such as first images uploaded by visual sensing devices and first point clouds uploaded by radar sensing devices. The first roadside device of this embodiment of the invention aggregates all the latest first images received from visual sensing devices at the same time or within the same hour to form an image sequence, i.e., a first image sequence; similarly, it aggregates all the latest first point clouds received from radar sensing devices at the same time or within the same hour to form a point cloud sequence, i.e., a first point cloud sequence. After obtaining the latest first image sequence, the first roadside device in this embodiment of the invention will perform image fusion and stitching based on the viewpoints and intrinsic and extrinsic parameters of each visual sensing device to obtain a panoramic image with the maximum viewpoint, namely the first panoramic image; similarly, after obtaining the latest first point cloud sequence, the first roadside device will perform point cloud coordinate calibration, deduplication and fusion based on the viewpoints and intrinsic and extrinsic parameters of each radar sensing device and stitch together a maximum point cloud set, namely the first panoramic point cloud.

[0080] It should be noted that the first roadside device in this embodiment of the invention can perform image fusion and stitching based on the viewpoints and intrinsic and extrinsic parameters of each visual sensing device in multiple ways. One implementation is to fuse the viewpoints and intrinsic and extrinsic parameters of all visual sensing devices (bullet cameras and fisheye cameras) corresponding to the first image sequence, and stitch all the first images in the first image sequence based on the fused viewpoints and fused camera intrinsic and extrinsic parameters to obtain the corresponding first panoramic image. Another implementation is to select the first images of one or more visual sensing devices of a specified camera type from the first image sequence and stitch them together to obtain the corresponding first panoramic image. Here, the specified camera type is usually set to a bullet camera. If the number of visual sensing devices of the specified camera type is unique, the first image of that device is directly used as the corresponding first panoramic image. Yet another implementation is to select the first images of one or more visual sensing devices with pre-specified device numbers from the first image sequence and stitch them together to obtain the corresponding first panoramic image. If the number of visual sensing devices with pre-specified device numbers is unique, the first image corresponding to that device is directly used as the corresponding first panoramic image.

[0081] It should also be noted that the first roadside device in this embodiment of the invention can perform point cloud coordinate calibration, deduplication, and fusion based on the viewpoint and intrinsic / extrinsic parameters of each radar sensing device in multiple ways. One implementation method is to fuse the viewpoint and intrinsic / extrinsic parameters of all radar sensing devices (millimeter-wave radar and lidar) corresponding to the first point cloud sequence, and perform point cloud coordinate calibration, deduplication, and fusion on all first point clouds in the first point cloud sequence based on the fused viewpoint and fused radar intrinsic / extrinsic / extrinsic parameters to obtain the corresponding first panoramic point cloud. Another implementation method is to select the first point clouds of one or more radar sensing devices of a specified radar type from the first point cloud sequence and perform point cloud coordinate calibration, deduplication, and fusion. The first panoramic point cloud is obtained by combining the radar types specified here, including millimeter-wave radar and lidar. The radar type can be dynamically set based on the needs of the actual scene. LiDAR is selected by default. If the number of radar sensing devices with the specified radar type is unique, the first point cloud of that device is directly used as the corresponding first panoramic point cloud. Another implementation method is to select the first point clouds of one or more radar sensing devices with pre-specified device numbers from the first point cloud sequence, perform point cloud coordinate calibration, deduplication, and fusion to obtain the corresponding first panoramic point cloud. If the number of radar sensing devices with the specified device numbers is unique, the first point cloud of that device is directly used as the corresponding first panoramic point cloud.

[0082] Step 2: Based on the first panoramic image, perform static target identification and target attribute identification on the road surface and update the static target tracking sequence set according to the identification results; and based on the first panoramic image, perform road surface condition target identification and update the road surface target tracking sequence set according to the identification results.

[0083] Specifically, this includes: Step 21, performing road static target identification and target attribute identification based on the first panoramic image, and updating the static target tracking sequence set based on the identification results;

[0084] The static target tracking sequence set includes multiple first target tracking sequences; each first target tracking sequence includes a first target number, a first target type, and multiple first target tracking data; the first target type includes intersections, roads, road boundaries, lanes, lane markings, poles, traffic signs, arrows, U-turn areas, stop lines, traffic lights, pedestrian crossings, no-entry zones within intersections, safety islands, left-turn waiting areas, traffic guidance areas, and obstacles; the first target tracking data includes first target coordinates, first target geometric parameters, first target attribute sets, and first timestamps; here, in this embodiment of the invention, the first roadside device uses the static target tracking sequence set to perform state tracking on various local static targets.

[0085] Specifically, this includes: step 211, performing road static target recognition and classification processing on the first panoramic image to obtain multiple corresponding first target recognition boxes; extracting the image area covered by each first target recognition box on the first panoramic image as the corresponding first target image; and using the time information corresponding to the first panoramic image as the corresponding first timestamp.

[0086] The first target recognition box (bbox) includes the coordinates of its center point, its geometric dimensions, and the target type. The target types include intersections, roads, road boundaries, lanes, lane markings, poles, traffic signs, arrows, U-turn areas, stop lines, traffic lights, pedestrian crossings, no-entry zones within intersections, safety islands, left-turn waiting areas, traffic dividers, and obstacles. The geometric dimensions refer to the length, width, and height of the recognition box.

[0087] Here, the first roadside device in this embodiment of the invention performs road static target recognition and classification processing on the first panoramic image based on a conventional visual image target recognition and classification model. Common visual image target recognition and classification models include classification models based on convolutional neural networks, YOLO models, etc.

[0088] Step 212, and based on the preset panoramic camera intrinsic and extrinsic parameters, perform a coordinate information transformation from pixel coordinate system to world coordinate system for the center point coordinates of the first recognition box of each first target recognition box; and based on the preset panoramic camera intrinsic and extrinsic parameters, perform a size information transformation from image size to real size for the geometric dimensions of the first recognition box of each first target recognition box.

[0089] Here, the intrinsic and extrinsic parameters of the panoramic camera are obtained by weighted fusion of the intrinsic and extrinsic parameters of all visual perception devices in advance;

[0090] Step 213: Traverse each first target recognition box; during traversal, the currently traversed first target recognition box is taken as the corresponding current recognition box; the first target tracking sequence in the static target tracking sequence set whose first target type matches the target type of the first recognition box of the current recognition box is recorded as the corresponding first pre-selected sequence; calculate the corresponding first straight-line distance between the first target coordinates of the last first target tracking data in each first pre-selected sequence and the center point coordinates of the first recognition box of the current recognition box, and take the shortest distance as the corresponding shortest straight-line distance; identify whether the shortest straight-line distance exceeds the preset first shortest distance threshold; if it does not exceed the threshold, take the first pre-selected sequence corresponding to the shortest straight-line distance as the corresponding first matching sequence; if it exceeds the threshold, then in the static target... A new first target tracking sequence is added to the tracking sequence set as the corresponding first matching sequence, and a unique number is assigned to the first matching sequence as the first target number of the first matching sequence. The target type of the first recognition box of the current recognition box is used as the first target type of the first matching sequence. After obtaining the first matching sequence, the center point coordinates, geometric dimensions, and timestamp of the first recognition box corresponding to the current recognition box are used as the corresponding first target coordinates, first target geometric parameters, and first timestamp. Static target attribute extraction processing is performed on the first target image corresponding to the current recognition box to obtain the corresponding first target attribute set. The obtained first target coordinates, first target geometric parameters, first target attribute set, and first timestamp are combined to form a corresponding first target tracking data and added to the first matching sequence.

[0091] Here, the first shortest distance threshold is a pre-set short-distance threshold parameter; in this embodiment of the invention, the first roadside device performs static target attribute extraction processing based on the first target image corresponding to the current recognition box, and performs customized attribute feature extraction according to different target types; for example: extracting the intersection type (T-junction, crossroad, straight intersection, etc.) attribute for intersection targets; extracting attributes such as road signs, road grade, road direction, number of road lanes, road length, road accident frequency level, and road coordinate range for road targets; and extracting attributes such as road boundary targets. The system extracts attributes such as road markings, boundary type, boundary length, and boundary coordinate range from the target; it extracts attributes such as road segmentation markings, lane markings, lane direction, speed range, speed limit, traffic restriction, tidal flow lanes, and lane address range from the lane target; it extracts attributes such as lane marking type, lane marking length, and lane marking sampling point address from the lane target; it extracts attributes such as pole type, pole color, pole occlusion status, pole damage status, and pole fading status from the pole target; and it extracts attributes such as sign type, sign color, and traffic sign type from the traffic sign target. The system extracts attributes such as sign shape, sign obstruction status, sign damage status, sign fading status, and sign legibility; extracts attributes such as arrow type, arrow color, arrow direction, and arrow clarity from ground arrow targets; extracts attributes such as U-turn method and permitted U-turn time from U-turn area targets; extracts attributes such as stop line type and stop line clarity from stop line targets; extracts attributes such as traffic light type, traffic light color, traffic light control direction, traffic light obstruction status, and traffic light operating status from traffic light targets; and extracts attributes such as pedestrian crossings. Extracting attributes such as direction and pedestrian crossing clarity; extracting attributes such as area shape and clarity for prohibited areas within intersections; extracting attributes such as safety island shape, color, and damage status for safety islands; extracting attributes such as shape, connecting lane markings, and clarity for left-turn waiting areas; extracting attributes such as shape, connecting lane markings, and clarity for traffic flow guiding areas; extracting attributes such as obstacle type, color, occlusion status, and movement status for obstacles.

[0092] It should be noted that each attribute may include one or more attribute values. For example, the attribute values ​​of the speed limit attribute include the first speed limit period and the first speed limit threshold; the attribute values ​​of the traffic restriction attribute include the first traffic restriction period and the first set of restricted vehicle types; the attribute values ​​of the tidal lane attribute include the first lane change period and the road travel direction after the first lane change; and the lane address range attribute is composed of the set of lane edge sampling point addresses.

[0093] Step 214: Confirm whether there are any newly added first target tracking sequences in the static target tracking sequence set during this processing. If so, take each newly added first target tracking sequence as the corresponding first newly added sequence; traverse all first newly added sequences; during traversal, take the currently traversed first newly added sequence as the corresponding current newly added sequence, and take the first target type of the current newly added sequence as the corresponding current target type; when the current target type is one of the preset multiple types of movable target types, record other first target tracking sequences in the static target tracking sequence set whose first target type matches the current target type but are not currently newly added sequences as the corresponding pre-selected similar sequences; when the number of pre-selected similar sequences is not 0, record the last first target tracking data of each pre-selected similar sequence as the corresponding pre-selected target data, and take the first first target tracking data of the currently newly added sequence as the corresponding current target data; calculate the difference ratio for each geometric item value in the first target geometric parameters of each pre-selected target data and the current target data to obtain the corresponding first geometric item difference ratio set; and then... Each attribute item in the first target attribute set of each pre-selected target data and the current target data is converted into an attribute value and a difference calculation is performed based on the numerical attribute to obtain the corresponding first attribute item difference value set; the pre-selected target data whose first geometric item difference ratios in the first geometric item difference ratio set do not exceed the preset difference ratio threshold and whose first attribute item difference values ​​in the first attribute item difference value set are all 0 are recorded as the corresponding matching target data; if the number of matching target data is 1, the current target data is added to the first target tracking sequence corresponding to the matching target data and the newly added sequence is deleted from the static target tracking sequence set; if the number of matching target data is greater than 1, the average of all first geometric item difference ratios in the first geometric item difference ratio set corresponding to each matching target data is calculated to obtain the corresponding first average ratio, and the matching target data corresponding to the smallest first average ratio is taken as the corresponding final matching target data, the current target data is added to the first target tracking sequence corresponding to the final matching target data, and the newly added sequence is deleted from the static target tracking sequence set;

[0094] The various types of movable targets include poles, traffic signs, traffic lights, safety islands, and obstacles; the first geometric item difference ratio set includes multiple first geometric item difference ratios, where the first geometric item difference ratio = |the geometric item value of the current target data - the geometric item value of the pre-selected target data| / the geometric item value of the pre-selected target data; the first attribute item difference value set includes multiple first attribute item difference values, where the first attribute item difference value = (the attribute item value of the current target data - the attribute item value of the pre-selected target data).

[0095] Step 22, and perform road surface condition target identification based on the first panoramic image and update the road surface target tracking sequence set based on the identification results;

[0096] The road surface target tracking sequence set includes multiple second target tracking sequences; each second target tracking sequence includes a second target number, a second target type, and multiple second target tracking data; the second target types include accident road surfaces, spilled road surfaces, waterlogged road surfaces, snow-covered road surfaces, icy road surfaces, cracked road surfaces, pothole road surfaces, collapsed road surfaces, and construction road surfaces; the second target tracking data includes second target coordinates, a set of vertex coordinates of the first target range, and a second timestamp; here, in this embodiment of the invention, the first roadside device uses the road surface target tracking sequence set to track the status of various conditions occurring on the local road surface;

[0097] Specifically, this includes: step 221, performing road condition target recognition and classification processing on the first panoramic image to obtain multiple corresponding second target recognition boxes; and using the time information corresponding to the first panoramic image as the corresponding second timestamp;

[0098] The second target recognition box includes the coordinates of the center point of the second recognition box, the geometric dimensions of the second recognition box, and the target type of the second recognition box; the target type of the second recognition box includes accident road surface, spilled road surface, waterlogged road surface, snow-covered road surface, icy road surface, cracked road surface, pothole road surface, collapsed road surface, and construction road surface; the geometric dimensions of the second recognition box refer to the length, width, height and other information of the recognition box;

[0099] Here, the first roadside device in this embodiment of the invention performs road condition target recognition on the first panoramic image based on a conventional visual image target recognition model. Common visual image target recognition models include target recognition models based on multilayer perceptual neural networks and target recognition models based on convolutional neural networks.

[0100] Step 222: Identify the road surface area of ​​the first panoramic image to obtain the corresponding first road surface area; take the intersection surface between each second target recognition box and the first road surface area as the corresponding first intersection surface; transform each first intersection surface into a convex polygon to obtain the corresponding first convex polygon; and extract all vertex coordinates of each first convex polygon to form the corresponding first convex polygon vertex coordinate set.

[0101] Here, the first roadside device in this embodiment of the invention identifies the road surface area of ​​the driving road based on a conventional image semantic segmentation model;

[0102] Step 223, and based on the preset panoramic camera intrinsic and extrinsic parameters, perform a coordinate information transformation from pixel coordinate system to world coordinate system for the center point coordinates of the second recognition box of each second target recognition box; and based on the preset panoramic camera intrinsic and extrinsic parameters, perform a coordinate information transformation from pixel coordinate system to world coordinate system for the vertex coordinates of the first convex polygon vertex coordinate set corresponding to each second target recognition box.

[0103] Step 224 involves iterating through each second target recognition box, and during the iteration, using the currently traversed second target recognition box as the corresponding current recognition box; recording the second target tracking sequence in the road target tracking sequence set whose second target type matches the target type of the current recognition box as the corresponding second pre-selected sequence; when the number of second pre-selected sequences is 0, adding a second target tracking sequence to the road target tracking sequence set as the corresponding second matching sequence, assigning a unique number to the second matching sequence as the second target number of the second matching sequence, and using the target type of the current recognition box as the second target type of the second matching sequence; when the number of second pre-selected sequences is greater than 0, calculating the straight-line distance between the second target coordinates of the last second target tracking data in each second pre-selected sequence and the center point coordinates of the current recognition box to obtain the corresponding... The second straight-line distance is determined, and the shortest distance among them is taken as the corresponding shortest straight-line distance. It is then identified whether the shortest straight-line distance exceeds a preset second shortest distance threshold. If it does not exceed the threshold, the second pre-selected sequence corresponding to the shortest straight-line distance is taken as the corresponding second matching sequence. If it exceeds the threshold, a new second target tracking sequence is added to the road target tracking sequence set as the corresponding second matching sequence, and a unique number is assigned to the second matching sequence as the second target number of the second matching sequence. The target type of the second recognition box of the current recognition box is taken as the second target type of the second matching sequence. After obtaining the second matching sequence, the coordinates of the center point of the second recognition box corresponding to the current recognition box, the set of coordinates of the first convex polygon vertex and the second timestamp are taken as the corresponding second target coordinates, and the set of coordinates of the first target range vertex and the second timestamp are combined to form a corresponding second target tracking data and added to the second matching sequence.

[0104] Here, the second shortest distance threshold is a pre-set short distance length parameter.

[0105] Step 3: Identify whether all road static targets have undergone state changes based on the static target tracking sequence set to generate the corresponding first changed target sequence; update the corresponding map elements in the first vector map based on the first changed target sequence to generate the corresponding first updated element sequence; publish local data based on the first updated element sequence; and synchronize cloud platform data based on the first updated element sequence and the first panoramic cloud.

[0106] Specifically, it includes: Step 31, identifying whether all road static targets have undergone state changes based on the static target tracking sequence set, and generating the corresponding first change target sequence;

[0107] The first change target sequence includes one or more first change targets; the first change target includes a third target number, a third target type, a first change type, coordinates before the first change, coordinates after the first change, geometric parameters before the first change, geometric parameters after the first change, attribute set before the first change, attribute set after the first change, and a third timestamp; the first change type includes addition type, deletion type, movement type, and in-situ change type.

[0108] Specifically, this includes: step 311, statistically analyzing the number of first target tracking data in each first target tracking sequence of the static target tracking sequence set to generate a corresponding first quantity; and recording the first target tracking sequence with a first quantity of 1 as the corresponding newly added sequence; and recording the first target tracking sequence with a first quantity greater than 1 as the corresponding existing sequence;

[0109] Here, the static target corresponding to the first target tracking sequence with only one first target tracking data is newly added during the current update of the first panoramic image. In other words, such a first target tracking sequence actually corresponds to a newly added target.

[0110] Step 312, and use the time information corresponding to the first panoramic image as the current timestamp; and identify whether the first timestamp of the last first target tracking data of each stock sequence is earlier than the current timestamp; if so, change the current stock sequence to the corresponding stagnant sequence; if not, change the current stock sequence to the corresponding continuous sequence.

[0111] Here, the so-called stagnant sequence is the first target tracking sequence that has not been updated at least during the update of the first panoramic image. The reason it has not been updated is because the corresponding target has disappeared from its original position. In other words, such a sequence actually corresponds to a disappeared target.

[0112] Step 313: Extract the first target coordinates from the last two first target tracking data of each continuous sequence as the corresponding first and second coordinates; identify whether the straight-line distance between the first and second coordinates exceeds the preset first shortest distance threshold; if so, change the current continuous sequence to the corresponding moving sequence.

[0113] Here, because the target corresponding to each first target tracking sequence is a static target, the coordinates of the target generally do not change before and after time. Once a significant change occurs, that is, the straight-line distance between the first and second coordinates exceeds the first shortest distance threshold, it means that the target has been moved. In other words, such a sequence actually corresponds to a target that has been moved.

[0114] Step 314, and extract the first timestamp of the second-to-last first target tracking data of any remaining continuous sequence as the corresponding reference timestamp; and identify whether the first timestamp of the last first target tracking data of each stagnant sequence is earlier than the reference timestamp; if so, the current stagnant sequence is re-recorded as the corresponding long-term stagnant sequence; if not, the current stagnant sequence is re-recorded as the corresponding current period stagnant sequence.

[0115] Here, the target corresponding to the long-term stagnation sequence was found to have disappeared at the previous moment or even earlier; the target corresponding to the current stagnation sequence was found to have disappeared at the current moment.

[0116] Step 315, and initialize the first target sequence of the change to an empty sequence;

[0117] Step 316: When the number of newly added sequences is not zero, traverse each newly added sequence; during traversal, take the currently traversed newly added sequence as the corresponding current sequence; set the corresponding third target number to the first target number of the current sequence, set the corresponding third target type to the first target type of the current sequence, set the corresponding first change type to the new type, set the corresponding first change coordinates before the change to empty, set the corresponding first change coordinates after the change to the first target coordinates of the first target tracking data of the current sequence, set the corresponding first change geometric parameters before the change to empty, set the corresponding first change geometric parameters after the change to the first target geometric parameters of the first target tracking data of the current sequence, set the corresponding first change attribute set before the change to empty, set the corresponding first change attribute set to the first target attribute set of the first target tracking data of the current sequence, and set the corresponding third timestamp to the first timestamp of the first target tracking data of the current sequence; and add the corresponding first change target sequence to the first change target sequence by combining the obtained third target number, third target type, first change type, first change coordinates before the change, first change coordinates after the change, first change geometric parameters before the change, first change geometric parameters after the change, first change attribute set before the change, first change attribute set after the change, and third timestamp.

[0118] Here, in this embodiment of the invention, the first roadside device will add the target information of all newly added targets obtained in this step, i.e., all the first changed targets obtained in the current step, to the sequence used to summarize the changed target information, i.e., the first changed target sequence;

[0119] Step 317: When the number of current stagnant sequences is not 0, iterate through each current stagnant sequence; during the iteration, take the currently iterated current stagnant sequence as the corresponding current sequence; set the corresponding third target number to the first target number of the current sequence, set the corresponding third target type to the first target type of the current sequence, set the corresponding first change type to deletion type, set the corresponding first change coordinates before the change to the first target coordinates of the last first target tracking data of the current sequence, set the corresponding first change coordinates to empty, and set the corresponding first change geometric parameters before the change to the last first target tracking data of the current sequence. The first target geometric parameters, the corresponding first changed geometric parameters are set to empty, the corresponding first changed attribute set is set to the first target attribute set of the last first target tracking data in the current sequence, the corresponding first changed attribute set is set to empty, and the corresponding third timestamp is set to the time information corresponding to the first panoramic image; and the obtained third target number, third target type, first change type, first changed coordinates, first changed coordinates, first changed geometric parameters, first changed geometric parameters, first changed attribute set, first changed attribute set, and third timestamp are used to form the corresponding first changed target and add it to the first changed target sequence;

[0120] Here, in this embodiment of the invention, the first roadside device will add the target information of all the disappeared targets confirmed in this step, i.e., all the first changed targets obtained in the current step, to the sequence used to summarize the changed target information, i.e., the first changed target sequence;

[0121] Step 318: When the number of movement sequences is not zero, traverse each movement sequence; during traversal, take the currently traversed movement sequence as the corresponding current sequence, and record the last two first target tracking data of the current sequence as the corresponding tracking data before and after the change; set the corresponding third target number as the first target number of the current sequence, set the corresponding third target type as the first target type of the current sequence, set the corresponding first change type as the movement type, set the corresponding first before change coordinates as the first target coordinates of the tracking data before the change, set the corresponding first after change coordinates as the first target coordinates of the tracking data after the change, and set the corresponding first before change geometric parameters as the tracking data before the change. The first target geometric parameters of the tracking data, the corresponding first changed geometric parameters are set as the first target geometric parameters of the tracked data after the change, the corresponding first attribute set before the change is set as the first target attribute set of the tracked data before the change, the corresponding first changed attribute set is set as the first target attribute set of the tracked data after the change, and the corresponding third timestamp is set as the first timestamp of the tracked data after the change; and the obtained third target number, third target type, first change type, first change coordinates before the change, first change coordinates after the change, first change geometric parameters, first change geometric parameters, first change attribute set, first change attribute set and third timestamp are used to form the corresponding first changed target and add it to the first changed target sequence;

[0122] Here, in this embodiment of the invention, the first roadside device will add the target information of all the moving targets confirmed in this step, i.e., all the first changed targets obtained in the current step, to the sequence used to summarize the changed target information in this step, i.e., the first changed target sequence;

[0123] Step 319 involves iterating through each persistent sequence; during iteration, the currently iterated persistent sequence is taken as the current sequence, and the last two first target tracking data points of the current sequence are recorded as the corresponding first and second tracking data points; the difference ratio is calculated for each geometric item value in the first target geometric parameters of the first and second tracking data to obtain the corresponding second geometric item difference ratio set; the attribute value conversion is performed on each attribute item in the first target attribute set of the first and second tracking data, and the difference calculation is performed based on the value conversion attribute to obtain the corresponding second attribute item difference value set; it is then identified whether all second geometric item difference ratios in the second geometric item difference ratio set do not exceed a preset difference ratio threshold. If so, the corresponding first identification result is set as no change in geometric parameters; otherwise, the corresponding first identification result is set as a change in geometric parameters; it is then identified whether all first attribute item difference values ​​in the second attribute item difference value set are all 0. If so, the corresponding second identification result is set as no change in attributes; otherwise, the corresponding second identification result is set as a change in attributes; and the first identification result is set as a change in geometric parameters or the second... When the identification result indicates an attribute change, the corresponding third target number is set to the first target number of the current sequence, the corresponding third target type is set to the first target type of the current sequence, the corresponding first change type is set to the in-situ change type, the corresponding first pre-change coordinates are set to the first target coordinates of the first tracking data, the corresponding first post-change coordinates are set to the first target coordinates of the second tracking data, the corresponding first pre-change geometric parameters are set to the first target geometric parameters of the first tracking data, the corresponding first post-change geometric parameters are set to the first target geometric parameters of the second tracking data, the corresponding first pre-change attribute set is set to the first target attribute set of the first tracking data, the corresponding first post-change attribute set is set to the first target attribute set of the second tracking data, and the corresponding third timestamp is set to the first timestamp of the second tracking data. The obtained third target number, third target type, first change type, first pre-change coordinates, first post-change coordinates, first pre-change geometric parameters, first post-change geometric parameters, first pre-change attribute set, first post-change attribute set, and third timestamp form the corresponding first changed target and add it to the first changed target sequence.

[0124] The second geometric item difference ratio set includes multiple second geometric item difference ratios, and the second geometric item difference ratio = |geometric item value of the second tracking data - geometric item value of the first tracking data| / geometric item value of the first tracking data; the second attribute item difference value set includes multiple second attribute item difference values, and the second attribute item difference value = (attribute item value of the second tracking data - attribute item value of the first tracking data).

[0125] Here, in this embodiment of the invention, the first roadside device will add the target information of all static targets whose geometric dimensions and / or target attributes have changed in this confirmation, i.e., all the first changed targets obtained in the current step, to the sequence used to summarize the changed target information, i.e., the first changed target sequence;

[0126] Step 32, and update the corresponding map elements in the first vector map according to the first change target sequence to generate the corresponding first update element sequence;

[0127] The first updated element sequence includes multiple first updated elements; the first updated element includes the second element number, the second element type, the first element update type, the coordinates before the first element update, the coordinates after the first element update, the geometric parameters before the first element update, the geometric parameters after the first element update, the attribute set before the first element update, the attribute set after the first element update, and the first element update timestamp.

[0128] Specifically, this includes: when the first change target sequence is not empty, initializing the first update element sequence to an empty sequence; traversing each first change target in the first change target sequence; during traversal, taking the currently traversed first change target as the corresponding current change target, taking the third target type of the current change target as the corresponding current target type, and taking the first change coordinates, first change geometric parameters, and first change attribute set of the current change target as the corresponding previous coordinates, previous geometric parameters, and previous attribute set; and by querying the first vector map, taking the first map element whose first element type matches the current target type and whose first element coordinates, first element geometric parameters, and first element attribute set match the corresponding previous coordinates, previous geometric parameters, and previous attribute set as the corresponding current map element; and when the current map element is not empty, setting the corresponding second element number to the first element number of the current map element, setting the corresponding second element type to the first element type of the current map element, setting the corresponding first element update type to the first change type of the current change target, setting the corresponding first element update coordinates to the first change coordinates of the current change target, and setting the corresponding first element... The updated coordinates are the first changed coordinates of the current changed target; the corresponding first element's pre-update geometric parameters are set to the first changed geometric parameters of the current changed target; the corresponding first element's post-update geometric parameters are set to the first changed geometric parameters of the current changed target; the corresponding first element's pre-update attribute set is set to the first changed attribute set of the current changed target; the corresponding first element's post-update attribute set is set to the first changed attribute set of the current changed target; and the corresponding first element's update timestamp is set to the third timestamp of the current changed target. A corresponding first updated element is then added to the first updated element sequence, composed of the obtained second element number, second element type, first element update type, first element pre-update coordinates, first element post-update coordinates, first element pre-update geometric parameters, first element post-update geometric parameters, first element pre-update attribute set, first element post-update attribute set, and first element update timestamp. When the element in this sequence is successfully added, the first element coordinates, first element geometric parameters, and first element attribute set corresponding to the current map element are updated based on the first changed coordinates, first changed geometric parameters, and first changed attribute set of the current changed target.

[0129] Here, the first roadside device in this embodiment of the invention finds the map elements that need to be changed in the first vector map according to the first change target sequence, and forms a first update element sequence by the map element number of each map element and the first change target sequence; and after obtaining the first update element sequence, it updates the coordinates, size and attributes of all map elements that need to be changed in the local map in real time.

[0130] Step 33, and publish local data based on the first updated element sequence;

[0131] Specifically, this includes: when the first updated element sequence is not empty, publishing the first updated element sequence to a locally preset roadside map element update message queue so that any traffic participant client connected to the current first roadside device can subscribe to the message;

[0132] Among them, traffic participant clients include in-vehicle driving systems of any type of motor vehicle, navigation devices, equipment or systems of any type of non-motor vehicle, and navigation clients, devices, equipment or systems of any type of individual user;

[0133] Here, the first roadside device in this embodiment of the invention will have multiple message queues pre-configured locally, and each message queue supports message publishing, subscription / push; the roadside map feature update message queue is one of them, which is used to publish the latest local map update information; each traffic participant client can obtain the latest first updated feature sequence from this queue through message subscription; in addition, the first roadside device in this embodiment of the invention can also push the latest first updated feature sequence in the queue to all nearby traffic participant clients through periodic push.

[0134] Step 34, and synchronize cloud platform data based on the first updated element sequence and the first panoramic cloud;

[0135] Specifically, it includes: Step 341, when the first updated element sequence is not empty, the first updated element in the first updated element sequence whose second element type is one of the preset multiple static core element types is recorded as the corresponding core element; and the first updated element in the first updated element sequence whose second element type does not match any of the preset multiple static core element types is recorded as the corresponding non-core element.

[0136] Here, the various static core element types in this embodiment of the invention are a pre-set subset of static target types, which can be customized according to actual application needs. For example, it can be set to consist of traffic lights, lanes, stop lines, waiting areas, etc. This embodiment of the invention stipulates that before uploading the element update information of the static core element type to the cloud platform, its corresponding point cloud information must also be extracted and uploaded together.

[0137] Step 342, and the corresponding first update element sequence is composed of all non-core elements;

[0138] Step 343: When the first panoramic view cloud is not empty, perform point cloud target recognition on the first panoramic view cloud to obtain multiple corresponding third target recognition boxes; and take the point cloud in each third target recognition box of the first panoramic view cloud as the corresponding first target point cloud;

[0139] The third target recognition box includes the coordinates of the center point of the third recognition box and the geometric dimensions of the third recognition box; the geometric dimensions of the third recognition box are actually the length, width and height of the third target recognition box.

[0140] Here, the first roadside device in this embodiment of the invention will use a conventional point cloud target recognition model to process the point cloud target recognition of the first panoramic view cloud;

[0141] Step 344 involves iterating through each core element; during the iteration, the currently iterated core element is taken as the corresponding current element, and the first changed coordinates of the current element are taken as the corresponding current element coordinates; the straight-line distance between the current element coordinates and the center point coordinates of each third recognition box is calculated to generate the corresponding second straight-line distance; the minimum value among all the obtained second straight-line distances is selected as the corresponding third straight-line distance; it is then determined whether the third straight-line distance exceeds the preset third shortest distance threshold. If it does not exceed the threshold, the first target point cloud corresponding to the third straight-line distance is taken as the corresponding first element updated point cloud; if it exceeds the threshold, the corresponding first element updated point cloud is set to empty; the current element and the corresponding first element updated point cloud are combined to form a corresponding second updated element; and at the end of the iteration, all the obtained second updated elements are combined to form a corresponding second updated element sequence.

[0142] Here, the third shortest distance threshold is a pre-set small distance threshold length;

[0143] Step 345: Extract the locally preset roadside device numbers as the corresponding first roadside device numbers; extract the locally preset intersection or road numbers corresponding to the first roadside device numbers as the corresponding first intersection / road numbers; use the time information of the first panoramic image as the corresponding first report timestamp; form a corresponding first sensing device type set by the device types of all visual sensing devices corresponding to the first panoramic image; form a corresponding first sensing device number set by the device numbers of all visual sensing devices corresponding to the first panoramic image; form a corresponding second sensing device type set by the device types of all radar sensing devices corresponding to the first panoramic cloud; form a corresponding second sensing device number set by the device numbers of all radar sensing devices corresponding to the first panoramic cloud; and extract the first target tracking sequences corresponding to each first update element in the first update element sequence from the static target tracking sequence set to form a corresponding first target tracking sequence set.

[0144] Step 345, and send the corresponding first cloud platform synchronous data report composed of the obtained first roadside device number, first intersection / road number, first map version of the first vector map, first panoramic view, first report timestamp, first sensing device type set, first sensing device number set, second sensing device type set, second sensing device number set, first target tracking sequence set, first update element sequence and second update element sequence to the first cloud platform.

[0145] Step 4: Based on the first vector map, road events are identified to generate a corresponding first event list, and local data is published based on the first event list; based on the road surface target tracking sequence set and the first vector map, road events are identified to generate a corresponding second event list, and local data is published and cloud platform data is synchronized based on the second event list.

[0146] Here, the first event list is actually a dynamic list of events that publishes lane information that is currently in speed limit, traffic restriction and tidal flow lanes;

[0147] Specifically, this includes: Step 41, identifying road events based on the first vector map to generate a corresponding first event list, and publishing local data based on the first event list;

[0148] Specifically, this includes: step 411, generating a corresponding first event list based on road event recognition using the first vector map;

[0149] The first event list includes multiple first event records; each first event record includes the first event name, first event lane number, first event effective time period, and first event parameters; the first event name includes speed limit events, traffic restriction events, and tidal flow lane events; when the first event name is a speed limit event, the corresponding first event effective time period is the speed limit period, and the corresponding first event parameter is the speed limit threshold; when the first event name is a traffic restriction event, the corresponding first event effective time period is the traffic restriction period, and the corresponding first event parameter is the set of restricted vehicle types; when the first event name is a tidal flow lane event, the corresponding first event effective time period is the lane change period, and the corresponding first event parameter is the road travel direction after the lane change;

[0150] Specifically, this includes: step 4111, recording the first map element in the first vector map whose first element type is lane as the corresponding first lane element; and initializing the first event list to empty;

[0151] Step 4112 involves identifying whether a speed limit attribute exists in the first element attribute set of each first lane element; if it exists, the corresponding first speed limit period and first speed limit threshold are extracted from the speed limit attribute; and it is also identified whether the current time is not earlier than the start time of the first speed limit period; if so, the corresponding first event name is set as a speed limit event, the corresponding first event lane number is set as the first element number of the current first lane element, the corresponding first event effective period is set as the corresponding first speed limit period, and the corresponding first event parameter is set as the first speed limit threshold. A corresponding first event record is then added to the first event list, consisting of the obtained first event name, first event lane number, first event effective period, and first event parameter.

[0152] Step 4113 involves identifying whether a traffic restriction attribute exists in the first element attribute set of each first lane element; if it exists, the corresponding first traffic restriction period and first restricted vehicle type set are extracted from the traffic restriction attribute; and it is also identified whether the current time is not earlier than the start time of the first traffic restriction period; if so, the corresponding first event name is set as a traffic restriction event, the corresponding first event lane number is set as the first element number of the current first lane element, the corresponding first event valid period is set as the corresponding first traffic restriction period, and the corresponding first event parameters are set as the first restricted vehicle type set. A corresponding first event record is then added to the first event list, consisting of the obtained first event name, first event lane number, first event valid period, and first event parameters.

[0153] Step 4114 involves identifying whether a tidal lane attribute exists in the first element attribute set of each first lane element; if it exists, the corresponding first lane change time period and the road driving direction after the first lane change are extracted from the tidal lane attribute; and whether the current time is not earlier than the start time of the first lane change time period is identified; if so, the corresponding first event name is set as tidal lane event, the corresponding first event lane number is set as the first element number of the current first lane element, the corresponding first event effective time period is set as the corresponding first lane change time period, and the corresponding first event parameter is set as the road driving direction after the first lane change. A corresponding first event record is then added to the first event list, consisting of the obtained first event name, first event lane number, first event effective time period, and first event parameter.

[0154] Step 412, and publish local data according to the first event list;

[0155] Specifically, this includes: when the first event list is not empty, publishing the first event list to a locally preset lane time-limited event message queue so that any traffic participant client connected to the current first roadside device can subscribe to the message;

[0156] Among them, traffic participant clients include in-vehicle driving systems of any type of motor vehicle, navigation devices, equipment or systems of any type of non-motor vehicle, and navigation clients, devices, equipment or systems of any type of individual user;

[0157] Here, as mentioned above, the first roadside device in this embodiment of the invention will have multiple message queues pre-configured locally, each message queue supporting message publishing, subscription / push; the lane time-limited event message queue is one of them, used to publish the latest speed limit, traffic restriction and tidal lane information; each traffic participant client can obtain the latest first event list from this queue through message subscription; in addition, the first roadside device in this embodiment of the invention can also push the latest first event list in the queue to all nearby traffic participant clients through periodic push.

[0158] Step 42: Based on the road target tracking sequence set and the first vector map, road event identification is performed to generate a corresponding second event list, and local data is published and cloud platform data is synchronized based on the second event list;

[0159] Specifically, this includes: step 421, and generating a corresponding second event list based on the road target tracking sequence set and the first vector map for road event identification;

[0160] The second event list includes multiple second event records. Each second event record includes the second event name, first event type, first event center coordinates, first event range vertex coordinate set, first event range number of lanes, and first event range lane number set. The second event name includes road accident events, road spill events, road water accumulation events, road snow accumulation events, road icing events, road crack events, road pothole events, road collapse events, and road construction events. The first event type includes new type, range reduction type, range expansion type, and end type. When the first event type is a new type, range reduction type, or range expansion type, the information consisting of the corresponding first event center coordinates, first event range vertex coordinate set, first event range number of lanes, and first event range lane number set is the latest set of real-time information corresponding to the current event. When the first event type is an end type, the information consisting of the corresponding first event center coordinates, first event range vertex coordinate set, first event range number of lanes, and first event range lane number set is the last set of historical information corresponding to the current event.

[0161] Specifically, this includes: Step 4211, when the road target tracking sequence set is not empty, counting the number of second target tracking data in each second target tracking sequence of the road target tracking sequence set to generate a corresponding third quantity; and recording the second target tracking sequence with a third quantity of 1 as the corresponding newly added sequence; and recording the second target tracking sequence with a third quantity greater than 1 as the corresponding existing sequence;

[0162] Here, newly added sequences correspond to newly discovered road surface conditions, while existing sequences correspond to previously discovered road surface conditions;

[0163] Step 4212, and use the time information corresponding to the first panoramic image as the current timestamp; and identify whether the second timestamp of the last second target tracking data of each stock sequence is earlier than the current timestamp; if so, change the current stock sequence to the corresponding stagnant sequence; if not, change the current stock sequence to the corresponding continuous sequence.

[0164] Here, the newly added sequence corresponds to the road surface conditions that were previously discovered but have now disappeared, while the persistent sequence corresponds to the road surface conditions that were previously discovered and still exist.

[0165] Step 4213: Extract the set of vertex coordinates of the first target range from the last two first target tracking data of each continuous sequence as the corresponding first and second vertex coordinate sets; estimate the area of ​​the convex polygon of the road surface based on the first and second vertex coordinate sets to obtain the corresponding first and second estimated areas; calculate the corresponding first area difference and first area change ratio based on the area change difference and area change ratio of the first and second estimated areas, where first area difference = second estimated area - first estimated area, and first area change ratio = first area difference / first estimated area; identify whether the absolute value of the first area change ratio exceeds the preset change ratio threshold; if it does not exceed the threshold, re-record the current continuous sequence as the corresponding range-no-change sequence; if it exceeds the threshold, re-record the current continuous sequence as the corresponding range-increasing sequence when the first area difference is greater than 0, and re-record the current continuous sequence as the corresponding range-decreasing sequence when the first area difference is less than 0.

[0166] Here, the current influence range of the road conditions corresponding to the sequence with increased range has increased compared to the previous one, while the current influence range of the road conditions corresponding to the sequence with decreased range has decreased compared to the previous one.

[0167] Step 4214, and initialize the second event list to an empty sequence;

[0168] Step 4215: When the number of newly added sequences is not zero, traverse each newly added sequence; during traversal, take the currently traversed newly added sequence as the corresponding current sequence, take the first second target tracking data of the current sequence as the corresponding current target tracking data, and take the second target type of the current sequence as the corresponding current target type; when the current target type is accident road surface, spilled road surface, waterlogged road surface, snowlogged road surface, icy road surface, cracked road surface, pothole road surface, collapsed road surface, or construction road surface, set the corresponding second event name as the corresponding road accident event, road spilled event, road waterlogged event, road snowlogged event, road icing event, road cracked event, road pothole event, road collapse event, or road construction event; set the corresponding first event type as the newly added type; set the corresponding first event center coordinates as the second target coordinates of the current target tracking data; and set the corresponding first event range vertex coordinate set as the current target. The system tracks the first target range vertex coordinate set of the tracking data; and records the first map element in the first vector map whose first feature type is lane and whose lane address range attribute in the first feature attribute set intersects with the road surface convex polygon area corresponding to the first event range vertex coordinate set as the corresponding first event lane element; it also counts the number of first event lane elements to generate the corresponding first lane number, and sets the corresponding first event range lane number as the corresponding first lane number; it uses the first feature numbers of all first event lane elements to form the corresponding first event lane number set, and sets the corresponding first event range lane number set as the corresponding first event lane number set; and adds a corresponding second event record to the second event list, composed of the obtained second event name, first event type, first event center coordinates, first event range vertex coordinate set, first event range lane number, and first event range lane number set.

[0169] Here, in this embodiment of the invention, the first roadside device will add all the relevant information of all newly discovered road conditions, i.e. all the second event records obtained in the current step, to the list used to summarize all road condition change information, i.e., the second event list;

[0170] Step 4216: When the number of range-shrinking sequences is not zero, traverse each range-shrinking sequence; during traversal, take the currently traversed range-shrinking sequence as the corresponding current sequence, take the last second target tracking data of the current sequence as the corresponding current target tracking data, and take the second target type of the current sequence as the corresponding current target type; when the current target type is accident road surface, spilled road surface, waterlogged road surface, snow-covered road surface, icy road surface, cracked road surface, pothole road surface, collapsed road surface, or construction road surface, set the corresponding second event name as the corresponding road accident event, road spilled event, road water accumulation event, road snow accumulation event, road icing event, road crack event, road pothole event, road collapse event, or road construction event; set the corresponding first event type as range reduction type; set the corresponding first event center coordinates as the second target coordinates of the current target tracking data; and set the corresponding first event range vertex coordinate set. This involves defining the first target range vertex coordinate set for the current target tracking data; and recording the first map element in the first vector map whose first feature type is lane and whose lane address range attribute in the first feature attribute set intersects with the convex polygon area of ​​the road surface corresponding to the first event range vertex coordinate set as the corresponding second event lane element; statistically analyzing the number of first event lane elements to generate the corresponding number of second lanes, and setting the corresponding number of first event range lanes as the corresponding number of second lanes; combining the first feature numbers of all second event lane elements to form the corresponding second event lane number set, and setting the corresponding first event range lane number set as the corresponding second event lane number set; and adding a corresponding second event record to the second event list, composed of the obtained second event name, first event type, first event center coordinates, first event range vertex coordinate set, first event range lane number, and first event range lane number set.

[0171] Here, in this embodiment of the invention, the first roadside device will add all relevant information about the existing road surface conditions with reduced impact range discovered in this instance, i.e., all the second event records obtained in the current step, to the list used to summarize all road surface condition change information in this instance, i.e., the second event list.

[0172] Step 4217: When the number of range-expanding sequences is not zero, traverse each range-expanding sequence; during traversal, take the currently traversed range-expanding sequence as the corresponding current sequence, take the last second target tracking data of the current sequence as the corresponding current target tracking data, and take the second target type of the current sequence as the corresponding current target type; when the current target type is accident road surface, spilled road surface, waterlogged road surface, snow-covered road surface, icy road surface, cracked road surface, pothole road surface, collapsed road surface, or construction road surface, set the corresponding second event name as the corresponding road accident event, road spilled event, road water accumulation event, road snow accumulation event, road icing event, road crack event, road pothole event, road collapse event, or road construction event; set the corresponding first event type as range expansion type; set the corresponding first event center coordinates as the second target coordinates of the current target tracking data; and set the corresponding first event range vertex coordinate set. This involves defining the first target range vertex coordinate set for the current target tracking data; and recording the first map element in the first vector map whose first feature type is lane and whose lane address range attribute in the first feature attribute set intersects with the convex polygon area of ​​the road surface corresponding to the first event range vertex coordinate set as the corresponding third event lane element; statistically analyzing the number of third event lane elements to generate the corresponding number of third lanes, and setting the corresponding number of lanes in the first event range as the corresponding number of third lanes; forming the corresponding third event lane number set from the first feature numbers of all third event lane elements, and setting the corresponding first event range lane number set as the corresponding third event lane number set; and adding a corresponding second event record to the second event list, composed of the obtained second event name, first event type, first event center coordinates, first event range vertex coordinate set, first event range lane number, and first event range lane number set.

[0173] Here, in this embodiment of the invention, the first roadside device will add all relevant information about the existing road surface conditions whose impact range has expanded this time, i.e., all the second event records obtained in the current step, to the list used to summarize all road surface condition change information this time, i.e., the second event list.

[0174] Step 4218: When the number of stalled sequences is not zero, traverse each stalled sequence; during traversal, take the currently traversed stalled sequence as the corresponding current sequence, take the last second target tracking data of the current sequence as the corresponding current target tracking data, and take the second target type of the current sequence as the corresponding current target type; when the current target type is accident road surface, spilled road surface, waterlogged road surface, snow-covered road surface, icy road surface, cracked road surface, pothole road surface, collapsed road surface, or construction road surface, set the corresponding second event name as the corresponding road accident event, road spilled event, road water accumulation event, road snow accumulation event, road icing event, road crack event, road pothole event, road collapse event, or road construction event; set the corresponding first event type as the end type; set the corresponding first event center coordinates as the second target coordinates of the current target tracking data; and set the corresponding first event range vertex coordinate set as the first target range vertex coordinates of the current target tracking data. The system sets up a set of four lanes for each event. It then assigns a lane as the first element in the first vector map, and records the first map element whose first element type is "lane" and whose lane address range attribute in the first element attribute set intersects with the convex polygon area of ​​the road surface corresponding to the vertex coordinate set of the first event range as the corresponding fourth event lane element. It also counts the number of fourth event lane elements to generate the corresponding number of fourth lanes, and sets the corresponding number of lanes in the first event range as the corresponding number of fourth lanes. Finally, it adds a corresponding second event record to the second event list, consisting of the obtained second event name, first event type, first event center coordinates, first event range vertex coordinate set, first event range lane number, and first event range lane number set. Upon successful addition of the record to the list, it removes the current sequence from the road surface target tracking sequence set.

[0175] Here, in this embodiment of the invention, the first roadside device will add the last set of historical information related to the historical road surface condition that has disappeared this time, i.e., all the second event records obtained in the current step, to the list used to summarize all road surface condition change information this time, i.e., the second event list;

[0176] Step 422, and perform local data publishing and cloud platform data synchronization according to the second event list;

[0177] Specifically, this includes: Step 4221, when the second event list is not empty, publishing the second event record named "Road Accident Event" in the second event list to a locally preset traffic accident event message queue for any traffic participant client connected to the current first roadside device to subscribe to the message; publishing the second event record named "Road Construction Event" in the second event list to a locally preset road construction event message queue for any traffic participant client connected to the current first roadside device to subscribe to the message; publishing the second event records named "Road Spillage Event," "Road Water Accumulation Event," "Road Snow Accumulation Event," and "Road Icing Event" in the second event list to a locally preset road safety event message queue for any traffic participant client connected to the current first roadside device to subscribe to the message; and publishing the second event records named "Road Crack Event," "Road Pothole Event," and "Road Collapse Event" in the second event list to a locally preset road defect event message queue for any traffic participant client connected to the current first roadside device to subscribe to the message.

[0178] Among them, traffic participant clients include in-vehicle driving systems of any type of motor vehicle, navigation devices, equipment or systems of any type of non-motor vehicle, and navigation clients, devices, equipment or systems of any type of individual user;

[0179] Here, as mentioned above, the first roadside device in this embodiment of the invention will have multiple message queues pre-configured locally, each message queue supporting message publishing, subscription / push; 1) The traffic accident event message queue is one of the pre-configured queues, used to publish the latest traffic accident information. Each traffic participant client can obtain the latest traffic accident second event record from this queue through message subscription. In addition, the first roadside device in this embodiment of the invention can also push the latest traffic accident second event record in the queue to all nearby traffic participant clients through periodic push; 2) The road construction event message queue is one of the pre-configured queues, used to publish the latest road construction information. Each traffic participant client can obtain the latest road construction second event record from this queue through message subscription. In addition, the first roadside device in this embodiment of the invention can also push the latest road construction second event record in the queue to all nearby traffic participant clients through periodic push; 3) The road safety event message queue is one of the pre-configured queues, used to publish the latest road safety hazards (etc.). Information such as road spills, water accumulation, snow accumulation, and icing can be obtained from the queue by each traffic participant client through message subscription. Additionally, the first roadside device in this embodiment can periodically push the latest second event records of road safety hazards (such as road spills, water accumulation, snow accumulation, and icing) from the queue to all nearby traffic participant clients. 4) The road defect event message queue is one of the pre-set queues used to publish the latest road defect information (such as road cracks, potholes, and road collapses). Each traffic participant client can obtain the latest second event records of road defects (such as road cracks, potholes, and road collapses) from the queue through message subscription. Additionally, the first roadside device in this embodiment can periodically push the latest second event records of road defects (such as road cracks, potholes, and road collapses) from the queue to all nearby traffic participant clients.

[0180] Step 4222: Extract the locally preset roadside device number as the corresponding second roadside device number; extract the locally preset intersection or road number corresponding to the second roadside device number as the corresponding second intersection / road number; use the time information of the first panoramic image as the corresponding second report timestamp; and send the corresponding second cloud platform synchronous data report composed of the obtained second roadside device number, second intersection / road number, first map version of the first vector map, first panoramic image, second report timestamp, road surface target tracking sequence set and second event list to the first cloud platform.

[0181] Figure 2This is a schematic diagram of an electronic device provided in Embodiment 2 of the present invention. This electronic device can be a terminal device or server implementing the methods of the aforementioned embodiments, or it can be a terminal device or server connected to the aforementioned terminal device or server implementing the methods of the aforementioned embodiments. Figure 2 As shown, the electronic device may include: a processor 301 (e.g., CPU), a memory 302, and a transceiver 303; the transceiver 303 is coupled to the processor 301, and the processor 301 controls the transmission and reception operations of the transceiver 303. The memory 302 may store various instructions for performing various processing functions and implementing the processing steps described in the foregoing embodiments. Preferably, the electronic device according to the embodiments of the present invention further includes: a power supply 304, a system bus 305, and a communication port 306. The system bus 305 is used to implement communication connections between components. The communication port 306 is used for communication between the electronic device and other peripherals. Figure 2 The system bus 305 mentioned can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This system bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 2 The symbol is represented by a single thick line, but this does not imply that there is only one bus or one type of bus. The communication interface is used to enable communication between the database access device and other devices (e.g., clients, read-write libraries, and read-only libraries). Memory may include random access memory (RAM) and may also include non-volatile memory, such as at least one disk drive.

[0182] The processors mentioned above can be general-purpose processors, including central processing units (CPUs), network processors (NPs), graphics processing units (GPUs), etc.; they can also be digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0183] It should be noted that the embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when run on a computer, cause the computer to perform the methods and processes provided in the above embodiments.

[0184] This invention also provides a chip for executing instructions, which is used to perform the processing steps described in the foregoing method embodiments.

[0185] This invention provides a method for processing map element data for roadside equipment, an electronic device, and a computer-readable storage medium. The roadside equipment tracks local static road targets (intersections, roads, road boundaries, lanes, lane markings, poles, traffic signs, arrows, U-turn areas, stop lines, traffic lights, pedestrian crossings, no-entry zones within intersections, safety islands, left-turn waiting areas, traffic guidance areas, obstacles, etc.) and road surface conditions (accidents, spills, water accumulation, snow accumulation, icing, cracks, potholes, collapses, construction, etc.) based on the static target tracking information. It then confirms whether map elements have changed based on the static target tracking information and verifies the changes. The system performs real-time map updates and message dissemination locally, synchronizing the updated information to the cloud platform. Based on road condition tracking information, it identifies road events (accidents, spills, water accumulation, snow accumulation, icing, cracks, potholes, collapses, road construction, etc.) in the current area, combining the identification results with map elements to output a corresponding event list for real-time message dissemination and remote cloud platform synchronization. Simultaneously, the roadside equipment identifies road events (timed speed limit events, timed traffic restriction events, tidal lane events, etc.) based on the local map, combining the identification results with map elements to output a corresponding event list for real-time message dissemination. This invention's roadside equipment can update the local map in real-time without relying on a cloud platform, synchronize the latest updated data to the cloud platform, and provide richer local traffic information to local traffic participants. This invention improves the timeliness of roadside map updates, enhances the personalization of vehicle-road cooperation, and reduces the data processing pressure on the cloud platform.

[0186] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0187] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0188] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A map feature data processing method of a roadside device, characterized by, The method comprises: The first roadside device pre-receives a roadside device monitoring range vector map issued by a first cloud platform as a corresponding first vector map and saves it; and at any time, receives a first image group issued by all visual perception devices to form a corresponding first image sequence, and receives a first point cloud issued by all radar perception devices to form a corresponding first point cloud sequence; and generates a corresponding first panoramic map by splicing a roadside panoramic map according to the latest first image sequence; and generates a corresponding first panoramic point cloud by fusing a roadside panoramic point cloud according to the latest first point cloud sequence; According to the first panoramic map, road static target recognition and target attribute recognition are performed, and a static target tracking sequence set is updated according to the recognition result; and according to the first panoramic map, road surface condition target recognition is performed, and a road surface target tracking sequence set is updated according to the recognition result; According to the static target tracking sequence set, whether all road static targets have state changes is identified to generate a corresponding first change target sequence; and according to the first change target sequence, corresponding map elements in the first vector map are updated to generate a corresponding first update element sequence; and local data is published according to the first update element sequence; and cloud platform data is synchronized according to the first update element sequence and the first panoramic point cloud; According to the first vector map, road event recognition is performed to generate a corresponding first event list, and local data is published according to the first event list; and according to the road surface target tracking sequence set and the first vector map, road surface event recognition is performed to generate a corresponding second event list, and local data is published and cloud platform data is synchronized according to the second event list; The static target tracking sequence set comprises a plurality of first target tracking sequences; the first target tracking sequence comprises a first target number, a first target type and a plurality of first target tracking data; the first target type comprises an intersection, a road, a road boundary, a lane, a lane marking, a pole, a traffic sign, an arrow, a U-turn area, a stop line, a signal light, a pedestrian crossing, an intersection forbidden area, a safety island, a left turn waiting area, a flow guide area and an obstacle; the first target tracking data comprises a first target coordinate, a first target geometric parameter, a first target attribute set and a first timestamp; The static target tracking sequence set is updated according to the first panoramic map, road static target recognition and target attribute recognition, and the recognition result, specifically comprising: The first panoramic image is subjected to road static target recognition and classification processing to obtain a plurality of first target recognition boxes corresponding thereto; image regions covered by each of the first target recognition boxes on the first panoramic image are extracted as corresponding first target images; and time information corresponding to the first panoramic image is taken as a corresponding first timestamp; the first target recognition box includes a first recognition box center point coordinate, a first recognition box geometric size, and a first recognition box target type; the first recognition box target type includes an intersection, a road, a road boundary, a lane, a lane marking, a pole-shaped object, a traffic sign, an arrow, a U-turn area, a stop line, a signal lamp, a pedestrian crossing, an intersection no-entry area, a safety island, a left-turn waiting area, a flow guide area, and an obstacle; and based on preset panoramic camera internal and external parameters, the first recognition box center point coordinate of each of the first target recognition boxes is subjected to coordinate information conversion from a pixel coordinate system to a world coordinate system; and based on preset panoramic camera internal and external parameters, the first recognition box geometric size of each of the first target recognition boxes is subjected to size information conversion from an image size to a real size; each of the first target recognition boxes is iterated; and during iteration, the first target recognition box being iterated at present is taken as a corresponding current recognition box; the first target tracking sequence in the static target tracking sequence set that matches the first recognition box target type of the current recognition box is recorded as a corresponding first preselected sequence; a straight-line distance between the first target coordinate of the last first target tracking data in each of the first preselected sequences and the first recognition box center point coordinate of the current recognition box is calculated to obtain a corresponding first straight-line distance, and the shortest straight-line distance among them is taken as a corresponding shortest straight-line distance; whether the shortest straight-line distance exceeds a preset first shortest distance threshold is identified; if not, the first preselected sequence corresponding to the shortest straight-line distance is taken as a corresponding first matching sequence; if so, a first target tracking sequence is newly added in the static target tracking sequence set as a corresponding first matching sequence, a unique number is assigned to the first matching sequence as a first target number of the first matching sequence, and the first recognition box target type of the current recognition box is taken as the first target type of the first matching sequence; after the first matching sequence is obtained, the first recognition box center point coordinate, the first recognition box geometric size, and the first timestamp of the current recognition box are taken as corresponding first target coordinates, first target geometric parameters, and a first timestamp; and the first target image corresponding to the current recognition box is subjected to static target attribute extraction processing to obtain a corresponding first target attribute set; and the first target coordinate, the first target geometric parameter, the first target attribute set, and the first timestamp obtained are used to form a corresponding first target tracking data to be added to the first matching sequence. confirm whether the first target tracking sequence is newly added in the current processing process; if yes, each first target tracking sequence newly added in the current processing process is taken as a corresponding first newly added sequence; all the first newly added sequences are traversed; in the traversal, the first newly added sequence currently traversed is taken as a corresponding current newly added sequence, and the first target type of the current newly added sequence is taken as a corresponding current target type; when the current target type is one of a plurality of preset movable target types, other first target tracking sequences in the static target tracking sequence set that match the first target type and the current target type and are not the current newly added sequence are recorded as corresponding preselected same type sequences; when the number of the preselected same type sequences is not 0, the last first target tracking data of each preselected same type sequence is recorded as corresponding preselected target data, and the first target tracking data of the current newly added sequence is taken as corresponding current target data; the first geometric term difference proportion set is obtained by performing difference proportion calculation on each geometric term value in the first target geometric parameter of each preselected target data and the current target data; the first attribute term difference value set is obtained by performing attribute numerical conversion on each attribute term in the first target attribute set of each preselected target data and the current target data and performing difference calculation based on the numerical attribute; the preselected target data for which each first geometric term difference proportion in the first geometric term difference proportion set does not exceed a preset difference proportion threshold and each first attribute term difference value in the first attribute term difference value set is 0 is recorded as corresponding matching target data; if the number of the matching target data is 1, the current target data is added to the first target tracking sequence corresponding to the matching target data, and the current newly added sequence is deleted from the static target tracking sequence set; if the number of the matching target data is greater than 1, the first geometric term difference proportion set of each matching target data is taken as a corresponding first geometric term difference proportion set, the first mean proportion is obtained by performing mean value calculation on all the first geometric term difference proportions in the first geometric term difference proportion set, the matching target data corresponding to the smallest first mean proportion is taken as corresponding final matching target data, the current target data is added to the first target tracking sequence corresponding to the final matching target data, and the current newly added sequence is deleted from the static target tracking sequence set; wherein the plurality of movable target types include a rod-shaped object, a traffic sign, a signal light, a safety island and an obstacle; the first geometric term difference proportion set includes a plurality of first geometric term difference proportions, and a first geometric term difference proportion = |geometric term value of current target data - geometric term value of preselected target data| / geometric term value of preselected target data;The first attribute item difference numerical value set includes a plurality of the first attribute item difference numerical values, and the first attribute item difference numerical value=(attribute item numerical value of the current target data-preselected attribute item numerical value of the target data).

2. The map element data processing method of the roadside device according to claim 1, wherein, the first vector map is a vector map of a roadside area corresponding to the first roadside device; the first vector map comprises a first map version and a plurality of first map elements; the first map element comprises a first element number, a first element type, a first element coordinate, a first element geometric parameter, and a first element attribute set; the first element type comprises an intersection, a road, a road boundary, a lane, a lane marking, a pole, a traffic sign, an arrow, a U-turn area, a stop line, a signal light, a pedestrian crossing, an in-intersection no-entry area, a safety island, a left-turn waiting area, a flow guide area, and an obstacle; the device type of the visual perception device comprises a gun camera and a fisheye camera; the device type of the radar perception device comprises a millimeter wave radar and a laser radar; the set of road surface target tracking sequences comprises a plurality of second target tracking sequences; the second target tracking sequence comprises a second target number, a second target type, and a plurality of second target tracking data; the second target type comprises an accident road surface, a spill road surface, a water accumulation road surface, a snow accumulation road surface, an icy road surface, a crack road surface, a pothole road surface, a collapse road surface, and a construction road surface; the second target tracking data comprises a second target coordinate, a set of first target range vertex coordinates, and a second timestamp; the first change target sequence comprises one or more first change targets; the first change target comprises a third target number, a third target type, a first change type, a first change before coordinate, a first change after coordinate, a first change before geometric parameter, a first change after geometric parameter, a first change before attribute set, a first change after attribute set, and a third timestamp; the first change type comprises an addition type, a deletion type, a movement type, and an in-place change type; the first update element sequence comprises a plurality of first update elements; the first update element comprises a second element number, a second element type, a first element update type, a first element update before coordinate, a first element update after coordinate, a first element update before geometric parameter, a first element update after geometric parameter, a first element update before attribute set, a first element update after attribute set, and a first element update timestamp; the first event list comprises a plurality of first event records; the first event record comprises a first event name, a first event lane number, a first event valid time period, and a first event parameter; the first event name comprises a speed limit event, a traffic limit event, and a tidal lane event; when the first event name is a speed limit event, the corresponding first event valid time period is a speed limit time period, and the corresponding first event parameter is a speed limit threshold; when the first event name is a traffic limit event, the corresponding first event valid time period is a traffic limit time period, and the corresponding first event parameter is a traffic limit vehicle type set; when the first event name is a tidal lane event, the corresponding first event valid time period is a lane changing time period, and the corresponding first event parameter is a post-lane changing road driving direction. The second event list includes a plurality of second event records; the second event record includes a second event name, a first event type, a first event center coordinate, a first event range vertex coordinate set, a first event range lane number, and a first event range lane number set; the second event name includes a road surface accident event, a road surface spill event, a road surface water event, a road surface snow event, a road surface icing event, a road surface crack event, a road surface pit event, a road surface collapse event, and a road surface construction event; the first event type includes an added type, a range reduced type, a range expanded type, and an end type; when the first event type is an added type, a range reduced type, or a range expanded type, the set of information composed of the corresponding first event center coordinate, the first event range vertex coordinate set, the first event range lane number, and the first event range lane number set is the latest set of real-time information corresponding to the current event; when the first event type is an end type, the set of information composed of the corresponding first event center coordinate, the first event range vertex coordinate set, the first event range lane number, and the first event range lane number set is the last set of historical information corresponding to the current event.

3. The map feature data processing method of a roadside device according to claim 2, characterized by, The road surface condition target recognition is performed according to the first panoramic image, and a road surface target tracking sequence set is updated according to an identification result, specifically including: The first panoramic image is subjected to road surface condition target recognition and classification processing to obtain a plurality of corresponding second target recognition boxes; and the time information corresponding to the first panoramic image is taken as a corresponding second time stamp; the second target recognition box includes a second recognition box center point coordinate, a second recognition box geometric size, and a second recognition box target type; the second recognition box target type includes an accident road surface, a spill road surface, a water accumulation road surface, a snow accumulation road surface, an icing road surface, a crack road surface, a pit road surface, a collapse road surface, and a construction road surface; The first panoramic image is subjected to driving road surface area recognition to obtain a corresponding first road surface area; the intersection of each second target recognition box with the first road surface area is taken as a corresponding first intersection; each first intersection is subjected to convex polygon conversion to obtain a corresponding first convex polygon; and all vertex coordinates of each first convex polygon are extracted to form a corresponding first convex polygon vertex coordinate set; Based on a preset panoramic image camera internal and external parameter, the second recognition box center point coordinate of each second target recognition box is subjected to coordinate information conversion from a pixel coordinate system to a world coordinate system; and based on a preset panoramic image camera internal and external parameter, each vertex coordinate of the first convex polygon vertex coordinate set corresponding to each second target recognition box is subjected to coordinate information conversion from a pixel coordinate system to a world coordinate system; The second target recognition frame is traversed, the second target recognition frame being currently traversed is taken as a corresponding current recognition frame, the second target tracking sequence in the road surface target tracking sequence set that matches the second target type of the second target recognition frame of the current recognition frame is recorded as a corresponding second preselected sequence, when the number of second preselected sequences is 0, a second target tracking sequence is added to the road surface target tracking sequence set as a corresponding second matching sequence, a unique number is assigned to the second matching sequence as the second target number of the second matching sequence, and the second target type of the second matching sequence is taken as the second target type of the second matching sequence, when the number of second preselected sequences is greater than 0, a straight line distance between the second target coordinates of the last second target tracking data in each second preselected sequence and the second recognition frame center point coordinates of the current recognition frame is calculated to obtain a corresponding second straight line distance, the shortest straight line distance is taken as a corresponding shortest straight line distance, and it is identified whether the shortest straight line distance exceeds a preset second shortest distance threshold, if not, the second preselected sequence corresponding to the shortest straight line distance is taken as the second matching sequence, if yes, a second target tracking sequence is added to the road surface target tracking sequence set as a corresponding second matching sequence, a unique number is assigned to the second matching sequence as the second target number of the second matching sequence, and the second target type of the current recognition frame is taken as the second target type of the second matching sequence, and after the second matching sequence is obtained, the second recognition frame center point coordinates of the current recognition frame, the first convex polygon vertex coordinate set and the second timestamp are taken as the second target coordinates, the first target range vertex coordinate set and the second timestamp to form a corresponding second target tracking data, which is added to the second matching sequence.

4. The map feature data processing method of a roadside device according to claim 2, characterized by, The first change target sequence is generated according to the static target tracking sequence set, and the first change target sequence is generated according to the static target tracking sequence set, and the first change target sequence is generated according to the static target tracking sequence set. The number of first target tracking data of each first target tracking sequence in the static target tracking sequence set is counted to generate a corresponding first number, the first target tracking sequence with the first number of 1 is recorded as a corresponding new sequence, and the first target tracking sequence with the first number greater than 1 is recorded as a corresponding inventory sequence. The time information corresponding to the first panoramic image is taken as a current timestamp, and it is identified whether the first timestamp of the last first target tracking data of each inventory sequence is earlier than the current timestamp, if yes, the current inventory sequence is recorded as a corresponding static sequence, if not, the current inventory sequence is recorded as a corresponding continuous sequence. and the first target coordinates of the last two first target tracking data of each of the continuous sequences are extracted as corresponding first and second coordinates; and it is identified whether the straight-line distance between the first and second coordinates exceeds a preset first minimum distance threshold; if yes, the current continuous sequence is recorded as a corresponding moving sequence; and the first timestamp of the second-to-last first target tracking data of any of the remaining continuous sequences is extracted as a corresponding reference timestamp; and it is identified whether the first timestamp of the last first target tracking data of each of the stationary sequences is earlier than the reference timestamp; if yes, the current stationary sequence is recorded as a corresponding long-term stationary sequence; if not, the current stationary sequence is recorded as a corresponding current stationary sequence; and the first change target sequence is initialized as an empty sequence; and when the number of the new sequences is not 0, each of the new sequences is traversed; and during the traversal, the new sequence being currently traversed is taken as a corresponding current sequence; and the corresponding third target number is set as the first target number of the current sequence, the corresponding third target type is set as the first target type of the current sequence, the corresponding first change type is set as a new type, the corresponding first change front coordinates are set as empty, the corresponding first change rear coordinates are set as the first target coordinates of the first first target tracking data of the current sequence, the corresponding first change front geometric parameters are set as empty, the corresponding first change rear geometric parameters are set as the first target geometric parameters of the first first target tracking data of the current sequence, the corresponding first change front attribute set is set as empty, the corresponding first change rear attribute set is set as the first target attribute set of the first first target tracking data of the current sequence, and the corresponding third timestamp is set as the first timestamp of the first first target tracking data of the current sequence; and the third target number, the third target type, the first change type, the first change front coordinates, the first change rear coordinates, the first change front geometric parameters, the first change rear geometric parameters, the first change front attribute set, the first change rear attribute set, and the third timestamp obtained are added to the first change target sequence as a corresponding first change target; and when the number of the current stagnation sequences is not 0, traversing each of the current stagnation sequences; and when traversing, taking the current traversed current stagnation sequence as a corresponding current sequence; and setting a corresponding third target number as the first target number of the current sequence, setting a corresponding third target type as the first target type of the current sequence, setting a corresponding first change type as a deletion type, setting a corresponding first change before coordinate as the first target coordinate of the last first target tracking data of the current sequence, setting a corresponding first change after coordinate as null, setting a corresponding first change before geometric parameter as the first target geometric parameter of the last first target tracking data of the current sequence, setting a corresponding first change after geometric parameter as null, setting a corresponding first change before attribute set as the first target attribute set of the last first target tracking data of the current sequence, setting a corresponding first change after attribute set as null, and setting a corresponding third timestamp as time information corresponding to the first panoramic picture; and adding a corresponding first change target composed of the third target number, the third target type, the first change type, the first change before coordinate, the first change after coordinate, the first change before geometric parameter, the first change after geometric parameter, the first change before attribute set, the first change after attribute set, and the third timestamp into the first change target sequence; and when the number of the movement sequences is not 0, traversing each of the movement sequences; and when traversing, taking the movement sequence currently traversed as a corresponding current sequence, and taking the last two first target tracking data of the current sequence as corresponding pre-change tracking data and post-change tracking data; and setting a corresponding third target number as the first target number of the current sequence, setting a corresponding third target type as the first target type of the current sequence, setting a corresponding first change type as a movement type, setting a corresponding first pre-change coordinate as the first target coordinate of the pre-change tracking data, setting a corresponding first post-change coordinate as the first target coordinate of the post-change tracking data, setting a corresponding first pre-change geometric parameter as the first target geometric parameter of the pre-change tracking data, setting a corresponding first post-change geometric parameter as the first target geometric parameter of the post-change tracking data, setting a corresponding first pre-change attribute set as the first target attribute set of the pre-change tracking data, setting a corresponding first post-change attribute set as the first target attribute set of the post-change tracking data, and setting a corresponding third timestamp as the first timestamp of the post-change tracking data; and adding a corresponding first change target to the first change target sequence by the obtained third target number, third target type, first change type, first pre-change coordinate, first post-change coordinate, first pre-change geometric parameter, first post-change geometric parameter, first pre-change attribute set, first post-change attribute set, and third timestamp; and each of the first target geometric parameters of the first and second tracking data is subjected to difference ratio calculation to obtain a corresponding second geometric term difference ratio set, and each of the first target attribute sets of the first and second tracking data is subjected to attribute numerical conversion and difference calculation based on the numerical attribute to obtain a corresponding second attribute term difference value set; and whether all second geometric term difference ratios of the second geometric term difference ratio set are less than a preset difference ratio threshold is identified, and if yes, a corresponding first identification result is set as no change in geometric parameters, and if not, the corresponding first identification result is set as change in geometric parameters; and whether all first attribute term difference values of the second attribute term difference value set are 0 is identified, and if yes, a corresponding second identification result is set as no change in attribute, and if not, the corresponding second identification result is set as change in attribute; and when the first identification result is change in geometric parameters or the second identification result is change in attribute, a corresponding third target number is set as the first target number of the current sequence, a corresponding third target type is set as the first target type of the current sequence, a corresponding first change type is set as a stationary change type, a corresponding first change before coordinate is set as the first target coordinate of the first tracking data, a corresponding first change after coordinate is set as the first target coordinate of the second tracking data, a corresponding first change before geometric parameter is set as the first target geometric parameter of the first tracking data, a corresponding first change after geometric parameter is set as the first target geometric parameter of the second tracking data, a corresponding first change before attribute set is set as the first target attribute set of the first tracking data, a corresponding first change after attribute set is set as the first target attribute set of the second tracking data, and a corresponding third timestamp is set as the first timestamp of the second tracking data, and a corresponding first change target is added to the first change target sequence by the third target number, the third target type, the first change type, the first change before coordinate, the first change after coordinate, the first change before geometric parameter, the first change after geometric parameter, the first change before attribute set, the first change after attribute set, and the third timestamp; wherein the second geometric term difference ratio set includes a plurality of second geometric term difference ratios, and a second geometric term difference ratio = |geometric term value of second tracking data - geometric term value of first tracking data| / geometric term value of first tracking data.The second attribute item differential value set includes a plurality of second attribute item differential values, and the second attribute item differential value=(the attribute item value of the second tracking data-the attribute item value of the first tracking data).

5. The map feature data processing method of a roadside device according to claim 2, characterized by, The first vector map is updated according to the first change target sequence to generate a corresponding first update element sequence, specifically comprising: when the first change target sequence is not empty, initializing the first update element sequence as an empty sequence; and traversing each first change target of the first change target sequence; and when traversing, taking the currently traversed first change target as a corresponding current change target, taking the third target type of the current change target as a corresponding current target type, taking the first change before coordinate, the first change before geometric parameter and the first change before attribute set of the current change target as a corresponding before coordinate, before geometric parameter and before attribute set; and through querying the first vector map, taking the first map element with the first element type matching the current target type and the first element coordinate, the first element geometric parameter and the first element attribute set respectively matching the corresponding before coordinate, before geometric parameter and before attribute set as a corresponding current map element; and when the current map element is not empty, setting the corresponding second element number as the first element number of the current map element, setting the corresponding second element type as the first element type of the current map element, setting the corresponding first element update type as the first change type of the current change target, setting the corresponding first element update before coordinate as the first change before coordinate of the current change target, setting the corresponding first element update after coordinate as the first change after coordinate of the current change target, setting the corresponding first element update before geometric parameter as the first change before geometric parameter of the current change target, setting the corresponding first element update after geometric parameter as the first change after geometric parameter of the current change target, setting the corresponding first element update before attribute set as the first change before attribute set of the current change target, setting the corresponding first element update after attribute set as the first change after attribute set of the current change target, setting the corresponding first element update timestamp as the third timestamp of the current change target; and adding a corresponding first update element composed of the obtained second element number, second element type, first element update type, first element update before coordinate, first element update after coordinate, first element update before geometric parameter, first element update after geometric parameter, first element update before attribute set, first element update after attribute set and first element update timestamp into the first update element sequence; and when the sequence element is added successfully, updating the first element coordinate, the first element geometric parameter and the first element attribute set corresponding to the current map element according to the first change after coordinate, the first change after geometric parameter and the first change after attribute set of the current change target.

6. The map feature data processing method of a roadside device according to claim 2, characterized by, The local data publishing according to the first update element sequence specifically comprises: When the first update element sequence is not empty, the first update element sequence is published to a locally preset roadside map element update message queue for message subscription by any traffic participant client connected with the current first roadside device; the traffic participant client includes a vehicle-mounted driving system of any type of motor vehicle, a travel navigation device, equipment or system of any type of non-motor vehicle, and a navigation client, device, equipment or system of any type of personal user.

7. The map feature data processing method of a roadside device according to claim 2, characterized by, The cloud platform data synchronization according to the first update element sequence and the first panoramic point cloud specifically includes: When the first update element sequence is not empty, the first update element in the first update element sequence whose second element type is one of a plurality of preset static core element types is recorded as a corresponding core element; and the first update element in the first update element sequence whose second element type does not match any type of the plurality of preset static core element types is recorded as a corresponding non-core element; And all the non-core elements form a corresponding first update element sequence; When the first panoramic point cloud is not empty, point cloud target recognition is performed on the first panoramic point cloud to obtain a plurality of third target recognition boxes; and the point cloud in each third target recognition box of the first panoramic point cloud is taken as a corresponding first target point cloud; the third target recognition box includes a third recognition box center point coordinate and a third recognition box geometric size; Each core element is traversed; when traversing, the core element being currently traversed is taken as a corresponding current element, and the first changed coordinate of the current element is taken as a corresponding current element coordinate; the straight line distance between the current element coordinate and each third recognition box center point coordinate is calculated to obtain a corresponding second straight line distance; the minimum value of all the obtained second straight line distances is selected as a corresponding third straight line distance; whether the third straight line distance exceeds a preset third minimum distance threshold is identified; if not, the first target point cloud corresponding to the third straight line distance is taken as a corresponding first element updated point cloud; if yes, the corresponding first element updated point cloud is set as empty; the current element and the corresponding first element updated point cloud form a corresponding second update element; and when the traversal ends, all the obtained second update elements form a corresponding second update element sequence; and the local preset roadside device number is extracted as a corresponding first roadside device number; and the local preset intersection or road number corresponding to the first roadside device number is extracted as a corresponding first intersection / road number; and the time information of the first panoramic map is taken as a corresponding first report timestamp; and the device type of all the visual perception devices corresponding to the first panoramic map constitutes a corresponding first perception device type set; and the device number of all the visual perception devices corresponding to the first panoramic map constitutes a corresponding first perception device number set; and the device type of all the radar perception devices corresponding to the first panoramic point cloud constitutes a corresponding second perception device type set; and the device number of all the radar perception devices corresponding to the first panoramic point cloud constitutes a corresponding second perception device number set; and the first target tracking sequence corresponding to each first update element in the first update element sequence in the static target tracking sequence set is extracted to constitute a corresponding first target tracking sequence set; and the first roadside device number, the first intersection / road number, the first map version of the first vector map, the first panoramic map, the first report timestamp, the first perception device type set, the first perception device number set, the second perception device type set, the second perception device number set, the first target tracking sequence set, the first update element sequence and the second update element sequence obtained are used to constitute a corresponding first cloud platform synchronization data report, which is sent to the first cloud platform.

8. The map feature data processing method of a roadside device according to claim 2, characterized by, The road event identification according to the first vector map generates a corresponding first event list, specifically including: The first map element of the first element type of lane in the first vector map is recorded as a corresponding first lane element; and the first event list is initialized as empty; and whether the speed limit attribute exists in the first element attribute set of each first lane element is identified; if yes, the corresponding first speed limit period and first speed limit threshold are extracted from the speed limit attribute; and whether the current time is not earlier than the start time of the first speed limit period is identified; if yes, the corresponding first event name is set as a speed limit event, the corresponding first event lane number is set as the first element number of the current first lane element, the corresponding first event valid period is set as the corresponding first speed limit period, the corresponding first event parameter is set as the first speed limit threshold, and a corresponding first event record composed of the first event name, the first event lane number, the first event valid period and the first event parameter is added to the first event list; identify whether there is a limited driving attribute in the first element attribute set of each first lane element; if so, extract a corresponding first limited driving time period and a first limited driving vehicle type set from the limited driving attribute; identify whether the current time is not earlier than the start time of the first limited driving time period; if so, set the corresponding first event name as a limited driving event, set the corresponding first event lane number as the first element number of the current first lane element, set the corresponding first event effective time period as the corresponding first limited driving time period, set the corresponding first event parameter as the first limited driving vehicle type set, and add a corresponding first event record composed of the obtained first event name, first event lane number, first event effective time period, and first event parameter to the first event list; identify whether there is a tidal lane attribute in the first element attribute set of each first lane element; if so, extract a corresponding first lane changing time period and a first post-lane changing road driving direction from the tidal lane attribute; identify whether the current time is not earlier than the start time of the first lane changing time period; if so, set the corresponding first event name as a tidal lane event, set the corresponding first event lane number as the first element number of the current first lane element, set the corresponding first event effective time period as the corresponding first lane changing time period, set the corresponding first event parameter as the first post-lane changing road driving direction, and add a corresponding first event record composed of the obtained first event name, first event lane number, first event effective time period, and first event parameter to the first event list.

9. The map feature data processing method of a roadside device according to claim 2, characterized by, The local data publishing according to the first event list specifically includes: When the first event list is not empty, publishing the first event list to a locally preset lane time-limited event message queue for message subscription by any traffic participant client connected to the current first roadside device; the traffic participant client includes a vehicle-mounted driving system of any type of motor vehicle, a driving navigation device, equipment or system of any type of non-motor vehicle, and a navigation client, device, equipment or system of any type of personal user.

10. The map feature data processing method of a roadside device according to claim 2, characterized by, The road event identification according to the road target tracking sequence set and the first vector map to generate a corresponding second event list specifically includes: When the road target tracking sequence set is not empty, counting the number of second target tracking data of each second target tracking sequence of the road target tracking sequence set to generate a corresponding third number; and recording the second target tracking sequence with the third number being 1 as a corresponding new sequence; and recording the second target tracking sequence with the third number being greater than 1 as a corresponding inventory sequence; When the road target tracking sequence set is not empty, counting the number of second target tracking data of each second target tracking sequence of the road target tracking sequence set to generate a corresponding third number; and recording the second target tracking sequence with the third number being 1 as a corresponding new sequence; and recording the second target tracking sequence with the third number being greater than 1 as a corresponding inventory sequence; And the first panoramic map corresponding to the time information as the current time stamp; and each of the last second target tracking data of the inventory sequence of the second timestamp whether earlier than the current time stamp is identified; if yes, the current inventory sequence is changed to the corresponding stagnation sequence; if not, the current inventory sequence is changed to the corresponding continuous sequence; And the last two first target tracking data of each of the continuous sequence of the first target range vertex coordinate set is extracted as the corresponding first, second vertex coordinate set; and the first, second vertex coordinate set is used to estimate the area of the road convex polygon to obtain the corresponding first, second estimated area; and the first, second estimated area is used to calculate the area change difference and the area change ratio to produce the corresponding first area difference and the first area change ratio, the first area difference = the second estimated area - the first estimated area, the first area change ratio = the first area difference / the first estimated area; and whether the absolute value of the first area change ratio exceeds the preset change ratio threshold is identified; if not, the current continuous sequence is changed to the corresponding range no change sequence; if it has exceeded, the current continuous sequence is changed to the corresponding range larger sequence when the first area difference is greater than 0, and the current continuous sequence is changed to the corresponding range smaller sequence when the first area difference is less than 0; And the second event list is initialized to an empty sequence; and when the number of the added sequences is not 0, traversing each of the added sequences; and when traversing, taking the currently traversed added sequence as a corresponding current sequence, taking the first second target tracking data of the current sequence as a corresponding current target tracking data, and taking the second target type of the current sequence as a corresponding current target type; and when the current target type is an accident road surface, a spilled road surface, a water accumulated road surface, a snow accumulated road surface, an icy road surface, a crack road surface, a pothole road surface, a collapsed road surface, or a construction road surface, setting a corresponding second event name as a corresponding road accident event, a road spilled event, a water accumulated road event, a snow accumulated road event, an icy road event, a crack road event, a pothole road event, a collapsed road event, or a construction road event; setting a corresponding first event type as an added type; setting a corresponding first event center coordinate as the second target coordinate of the current target tracking data; setting a corresponding first event range vertex coordinate set as the first target range vertex coordinate set of the current target tracking data; recording, as a corresponding first event lane element, a first map element corresponding to a road area in which a lane address range property in the first element attribute set of the first element type of lane in the first vector map and a road convex polygon area corresponding to the first event range vertex coordinate set intersect; counting a number of the first event lane elements to generate a corresponding first lane number, and setting a corresponding first event range lane number as the corresponding first lane number; composing a corresponding first event lane number set from the first element numbers of all the first event lane elements, and setting a corresponding first event range lane number set as the corresponding first event lane number set; and adding a corresponding second event record composed of the obtained second event name, first event type, first event center coordinate, first event range vertex coordinate set, first event range lane number, and first event range lane number set to the second event list. and when the number of the range-reducing sequences is not 0, traversing each of the range-reducing sequences; and when traversing, taking the currently traversed range-reducing sequence as a corresponding current sequence, taking the last second target tracking data of the current sequence as a corresponding current target tracking data, and taking the second target type of the current sequence as a corresponding current target type; and when the current target type is an accident road surface, a spilled road surface, a water-accumulated road surface, a snow-accumulated road surface, an icy road surface, a cracked road surface, a pothole road surface, a collapsed road surface, or a construction road surface, setting a corresponding second event name as a corresponding road surface accident event, a road surface spilled event, a water-accumulated road surface event, a snow-accumulated road surface event, an icy road surface event, a cracked road surface event, a pothole road surface event, a collapsed road surface event, or a construction road surface event; setting a corresponding first event type as a range-reducing type; setting a corresponding first event center coordinate as the second target coordinate of the current target tracking data; setting a corresponding first event range vertex coordinate set as the first target range vertex coordinate set of the current target tracking data; and recording, as a corresponding second event lane element, a first map element corresponding to a road surface area in the first vector map, where the first element type is a lane and a lane address range attribute in the first element attribute set, and a road surface convex polygon area corresponding to the first event range vertex coordinate set; counting the number of the first event lane elements to generate a corresponding second lane number, and setting a corresponding first event range lane number as the second lane number; and setting a corresponding first event range lane number set as a corresponding second event lane number set composed of the first element numbers of all the second event lane elements; and adding, to the second event list, a corresponding second event record composed of the second event name, the first event type, the first event center coordinate, the first event range vertex coordinate set, the first event range lane number, and the first event range lane number set. and when the number of the range-increasing sequences is not 0, traversing each of the range-increasing sequences; and when traversing, taking the currently traversed range-increasing sequence as a corresponding current sequence, taking the last second target tracking data of the current sequence as a corresponding current target tracking data, and taking the second target type of the current sequence as a corresponding current target type; and when the current target type is an accident road surface, a spilled road surface, a water-accumulated road surface, a snow-accumulated road surface, an icy road surface, a cracked road surface, a pothole road surface, a collapsed road surface, or a construction road surface, setting a corresponding second event name as a corresponding road surface accident event, a road surface spilled event, a water-accumulated road surface event, a snow-accumulated road surface event, an icy road surface event, a cracked road surface event, a pothole road surface event, a collapsed road surface event, or a construction road surface event; setting a corresponding first event type as an expansion type; setting a corresponding first event center coordinate as the second target coordinate of the current target tracking data; setting a corresponding first event range vertex coordinate set as the first target range vertex coordinate set of the current target tracking data; recording, as a corresponding third event lane element, a first map element corresponding to a road surface area in the first vector map, where the first element type is a lane and a lane address range attribute in the first element attribute set, and a road surface convex polygon area corresponding to the first event range vertex coordinate set; counting the number of the third event lane elements to generate a corresponding third lane number, and setting a corresponding first event range lane number as the third lane number; composing a corresponding third event lane number set from the first element numbers of all the third event lane elements, and setting a corresponding first event range lane number set as the third event lane number set; and adding a corresponding second event record composed of the second event name, the first event type, the first event center coordinate, the first event range vertex coordinate set, the first event range lane number, and the first event range lane number set to the second event list. And when the number of the stagnation sequences is not 0, each of the stagnation sequences is traversed; and when traversing, the stagnation sequence currently traversed is taken as a corresponding current sequence, the last second target tracking data of the current sequence is taken as a corresponding current target tracking data, and the second target type of the current sequence is taken as a corresponding current target type; and when the current target type is an accident road surface, a spilled road surface, a water accumulated road surface, a snow accumulated road surface, an icy road surface, a crack road surface, a pothole road surface, a collapsed road surface or a construction road surface, a corresponding second event name is set as a corresponding road surface accident event, a road surface spilled event, a water accumulated road surface event, a snow accumulated road surface event, an icy road surface event, a crack road surface event, a pothole road surface event, a collapsed road surface event or a construction road surface event; a corresponding first event type is set as an end type; a corresponding first event center coordinate is set as the second target coordinate of the current target tracking data; a corresponding first event range vertex coordinate set is set as the first target range vertex coordinate set of the current target tracking data; the first map element corresponding to the road surface area intersected by the first event range vertex coordinate set and the road surface convex polygon area corresponding to the first element type of lane and the lane address range attribute in the first element attribute set in the first vector map is recorded as a corresponding fourth event lane element; a corresponding fourth lane number is generated by counting the number of the fourth event lane elements, and a corresponding first event range lane number is set as the fourth lane number; a corresponding fourth event lane number set is composed of the first element numbers of all the fourth event lane elements, and a corresponding first event range lane number set is set as the fourth event lane number set; a corresponding second event record is composed of the second event name, the first event type, the first event center coordinate, the first event range vertex coordinate set, the first event range lane number and the first event range lane number set, and is added to the second event list; and when the list record is added successfully, the current sequence is deleted from the road surface target tracking sequence set. 11.The map element data processing method of a roadside device according to claim 2, characterized by, The local data publishing and cloud platform data synchronization according to the second event list specifically include: when the second event list is not empty, publishing the second event record in the second event list with the second event name as a road surface accident event to a locally preset traffic accident event message queue for message subscription by any traffic participant client connected to the current first road side device; and publishing the second event record in the second event list with the second event name as a road surface construction event to a locally preset road construction event message queue for message subscription by any traffic participant client connected to the current first road side device; and publishing the second event record in the second event list with the second event name as a road surface spill event, a road surface water event, a road surface snow event and a road surface ice event to a locally preset road safety event message queue for message subscription by any traffic participant client connected to the current first road side device; and publishing the second event record in the second event list with the second event name as a road surface crack event, a road surface pothole event and a road surface collapse event to a locally preset road disease event message queue for message subscription by any traffic participant client connected to the current first road side device; the traffic participant client including a vehicle-mounted driving system of any type of motor vehicle, a travel navigation device, equipment or system of any type of non-motor vehicle, and a navigation client, device, equipment or system of any type of personal user; extracting a locally preset road side device number as a corresponding second road side device number; and extracting a locally preset intersection or road number corresponding to the second road side device number as a corresponding second intersection / road number; and extracting time information of the first panoramic map as a corresponding second report timestamp; and sending a corresponding second cloud platform synchronization data report to the first cloud platform by the obtained second road side device number, the second intersection / road number, the first map version of the first vector map, the first panoramic map, the second report timestamp, the road target tracking sequence set and the second event list.

12. An electronic device, comprising: comprising: a memory, a processor and a transceiver; the processor is configured to read and execute instructions in the memory to implement the method of any one of claims 1-11; the transceiver is coupled with the processor and controlled by the processor to perform message transmission and reception.

13. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, when the computer instructions are executed by a computer, the computer executes the method of any one of claims 1-11. The computer readable storage medium stores computer instructions, when the computer instructions are executed by a computer, the computer executes the method of any one of claims 1-11.

Images