Park remote monitoring system based on three-dimensional digitization

Abstract

The invention relates to the technical field of park video monitoring, in particular to a park remote monitoring system based on three-dimensional digitization, which comprises a video acquisition module for dividing a monitoring area in a park into a plurality of monitoring areas and arranging monitoring points in each monitoring area; collecting the video stream of each monitoring area through the monitoring point to obtain a plurality of area video streams; the video storage module registers the monitoring end corresponding to the monitoring point, establishes a communication channel between the monitoring end and the management end, and stores the video streams of each area to the management end based on the communication channel; the monitoring and early warning module is used for constructing a park three-dimensional model based on the plurality of regional video streams and carrying out risk assessment on the regional video streams based on the park three-dimensional model to obtain a risk assessment result, distribution of video management resources can be adjusted in time according to the actual situation of the monitored region and occurrence of abnormal events, the park monitoring video transmission efficiency is improved, and the safety of the park is improved. Therefore, the monitoring efficiency and the management level of the park are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of park video surveillance, and particularly to a park remote surveillance system based on three-dimensional digitization. Background Art

[0002] With the continuous expansion of the scale of the park and the increasing improvement of security requirements, it has become crucial to conduct efficient and accurate remote surveillance of the park. Traditional park surveillance systems often have many limitations: lacking a mechanism for dynamically adjusting the communication frequency, being unable to timely adjust the allocation of video management resources according to the actual situation of the surveillance area and the occurrence of abnormal events, thus affecting the efficiency of park surveillance video transmission, resulting in insufficient timeliness of the park surveillance efficiency and affecting the surveillance effect of the park.

[0003] Therefore, we propose a park remote surveillance system based on three-dimensional digitization to solve the above problems. Summary of the Invention

[0004] The purpose of the present invention is to provide a park remote surveillance system based on three-dimensional digitization to solve the problems raised in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A park remote surveillance system based on three-dimensional digitization, including:

[0006] A video acquisition module, which divides the surveillance areas in the park into multiple surveillance areas and arranges surveillance points in each surveillance area; collects video streams of each surveillance area through the surveillance points to obtain multiple area video streams;

[0007] A video storage module, which registers a monitoring end corresponding to the surveillance point, establishes a communication channel between the monitoring end and the management end, and stores each area video stream to the management end based on the communication channel;

[0008] A monitoring and warning module, which constructs a three-dimensional model of the park based on multiple area video streams, conducts a risk assessment on the area video streams based on the three-dimensional model of the park to obtain a risk assessment result, and issues a warning for the risk level and event scope of the surveillance area corresponding to the risk assessment result exceeding the preset conditions;

[0009] A frequency adjustment module, which obtains a sub-database having a mapping relationship with the abnormal surveillance area, obtains a communication channel between the abnormal surveillance area and the sub-database, and adjusts the transmission frequency of the communication channel between the abnormal surveillance area and the sub-database to obtain a target transmission frequency.

[0010] Preferably, the step of dividing the surveillance areas in the park into multiple surveillance areas and arranging surveillance points in each surveillance area includes:

[0011] Obtain the information of the park area to be monitored, where the park area information includes area function information and area change cycle;

[0012] Based on the area function information, divide the park into multiple monitoring areas, and based on the area change cycle, perform secondary area division on each monitoring area to obtain multiple sub-monitoring areas, where the sub-monitoring areas include primary monitoring areas and secondary monitoring areas;

[0013] Deploy monitoring points in each monitoring area.

[0014] Preferably, the steps of registering a monitoring terminal for the corresponding monitoring point, establishing a communication channel between the monitoring terminal and the management terminal, and storing each area video stream to the management terminal based on the communication channel include:

[0015] Set up multiple sub-databases in the management terminal and establish a mapping relationship between the sub-databases and the monitoring areas;

[0016] Determine the target video stream to be transmitted, and determine the sub-database corresponding to the target video stream according to the mapping relationship;

[0017] Establish a communication channel between the monitoring terminal and the sub-database, and obtain the transmission frequency corresponding to the communication channel, where the transmission frequency includes an initial transmission frequency or a target transmission frequency;

[0018] Transmit the target video stream to the corresponding sub-database based on the transmission frequency, where the target video stream is the first target video stream or the second target video stream;

[0019] Extract the initial network points in the first target video stream corresponding to the active network points of the second target video stream, and replace the initial network points in the first target video stream with the active network points of the second target video stream to obtain the regional target video stream at the latest time node;

[0020] Store each regional target video stream in the corresponding sub-database, and perform timestamp synchronization encoding on multiple area video streams at the same time node.

[0021] Preferably, the steps of determining the target video stream to be transmitted include:

[0022] Determine the target monitoring area, mark the actual feature points in the target monitoring area, mark virtual feature points corresponding to the actual feature points in the video stream, and connect the multiple virtual feature points to form a virtual feature network;

[0023] Obtain the area video stream corresponding to the initial virtual feature network as the first target video stream to be transmitted;

[0024] Obtain the initial virtual feature network by obtaining the corresponding virtual feature network in the first target video stream;

[0025] Identify active grid points from the initial virtual feature network to obtain a target grid block, and use the local video stream corresponding to the target grid block as the second target video stream to be transmitted.

[0026] Preferably, the monitoring and early warning module includes:

[0027] A model construction unit, configured to extract spatio-temporally aligned regional video streams from multiple sub-databases, and construct a three-dimensional model of the park based on the regional video streams;

[0028] A risk assessment unit, configured to extract abnormal event information in the three-dimensional model of the park, and evaluate the risk index of the corresponding abnormal event according to the abnormal event information;

[0029] An abnormal early warning unit, configured to obtain regional information of abnormal events corresponding to risk indices exceeding a preset index threshold according to the risk index of the abnormal event, where the regional information includes the abnormal event range and the abnormal risk index; use the monitoring area corresponding to the abnormal event range as the abnormal monitoring area and give an early warning.

[0030] Preferably, the steps of constructing a three-dimensional model of the park based on the regional video streams include:

[0031] Extract spatio-temporally aligned regional video streams from multiple sub-databases respectively, perform regional boundary marking on the monitoring areas corresponding to the multiple regional video streams, and assign a unique identifier to the regional boundary of each monitoring area to obtain a boundary identifier;

[0032] Fuse the multiple regional video streams according to the unique identifier to generate a park video stream;

[0033] Extract the three-dimensional point cloud in the park video stream, associate the boundary identifier in the regional video stream with the three-dimensional point cloud, and generate a primary three-dimensional model of the park;

[0034] Extract texture information from the video stream and fit it to the surface of the primary three-dimensional model of the park to obtain the three-dimensional model of the park.

[0035] Preferably, the steps of adjusting the transmission frequency of the communication channel between the abnormal monitoring area and the sub-database include:

[0036] Obtain the communication channel between the abnormal monitoring area and the sub-database to obtain a limited channel, and determine the target transmission frequency corresponding to each limited channel;

[0037] Obtain multiple independent communication channels, extract multiple limited channels caused by the same abnormal event from the multiple communication channels to obtain a channel group, activate all the limited channels in the channel group at the same time, and start the transmission of the target video stream;

[0038] Set the transmission frequencies of all the defined channels within the channel group to the same target transmission frequency.

[0039] Preferably, the step of determining the target transmission frequency corresponding to each defined channel includes:

[0040] Classify the abnormal risk index into multiple risk levels, and set the transmission frequency corresponding to the risk level.

[0041] Obtain the abnormal risk index of the abnormal monitoring area, and obtain the transmission frequency corresponding to the risk level to which the abnormal risk index belongs as the target transmission frequency of the corresponding defined channel.

[0042] Compared with the prior art, the beneficial effects of the present invention are:

[0043] 1. By obtaining the park area information, including area function information and area change cycle, based on the area function information, the park is divided into multiple monitoring areas, and further based on the area change cycle, each monitoring area is sub-divided into multiple sub-monitoring areas, and then monitoring points are reasonably arranged within each monitoring area. The refined division and arrangement method can ensure that the monitoring points can more effectively cover key areas, reduce monitoring blind spots, and improve the comprehensiveness and pertinence of monitoring.

[0044] 2. Set multiple sub-databases at the management end, and establish a mapping relationship between the sub-databases and the monitoring areas. By determining the target video stream to be transmitted, it is transmitted to the corresponding sub-database according to the mapping relationship, and timestamp synchronization coding is performed on the multiple area video streams at the same time node. It can make the management of video data more orderly, facilitate the quick and accurate retrieval and call of video data in specific areas, and improve the monitoring efficiency and management level.

[0045] 3. Obtain the sub-database that has a mapping relationship with the abnormal monitoring area, obtain the communication channel between the abnormal monitoring area and the sub-database, adjust the transmission frequency of the communication channel between the abnormal monitoring area and the sub-database to obtain the target transmission frequency, and set the transmission frequencies of all the defined channels within the channel group to the same target transmission frequency. The way of dynamically adjusting the communication frequency can ensure that the video transmission in important areas is guaranteed by sufficient communication resources, and at the same time avoid unnecessary waste of communication resources, and improve the utilization efficiency of communication resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings without creative efforts.

[0047] Figure 1 This is the system structure block diagram of the present invention. Specific embodiments

[0048] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0049] Embodiment

[0050] Please refer to Figure 1 , the present invention provides a technical solution for a park remote monitoring system based on three-dimensional digitization: A park remote monitoring system based on three-dimensional digitization includes:

[0051] A video acquisition module divides the monitoring areas in the park into multiple monitoring areas, and arranges monitoring points in each monitoring area; multiple area video streams are obtained by collecting the video streams of each monitoring area through the monitoring points;

[0052] The steps of dividing the monitoring areas in the park into multiple monitoring areas and arranging monitoring points in each monitoring area include: obtaining the park area information to be monitored, where the park area information includes area function information and area change cycle; based on the area function information, the park is divided into multiple monitoring areas, and each monitoring area is secondarily divided into multiple sub-monitoring areas based on the area change cycle, where the sub-monitoring areas include primary monitoring areas and secondary monitoring areas; monitoring points are arranged in each monitoring area;

[0053] Specifically, obtain the regional information within the park, such as data on the planning drawings, building layout, and functional zoning of the park, so as to determine the overall architecture of the park and the basic uses of each area. According to the regional function information, such as the office management area, the living service area, the production operation area, etc., conduct a preliminary regional division to form multiple monitoring areas. Each area of a functional type corresponds to a monitoring area. The monitoring area is further divided to obtain sub-monitoring areas, and the sub-monitoring areas are classified into primary monitoring areas and secondary monitoring areas. The sub-monitoring areas whose regional change cycle exceeds the preset cycle time are used as primary monitoring areas and are used for subsequent video stream collection according to the regional change cycle. The sub-monitoring areas whose regional change cycle does not exceed the preset cycle time are used as secondary monitoring areas. The regional change cycle refers to the number of changes and the change time of each feature point within the area. When the number of changing feature points exceeds the preset threshold of the overall feature points, calculate the change time of the overall changing feature points. If the change time exceeds the preset cycle time, the corresponding sub-area of the area is determined as a primary monitoring area; otherwise, it is determined as a secondary monitoring area. Among them, the feature points and the descriptors corresponding to the feature points can be calibrated according to the situation within the sub-monitoring area, such as the location of the facilities and the placement direction of the facilities, or the building information within the area, such as the building shape for feature point marking. The construction of buildings and the placement change cycle of facilities are generally long, so the corresponding areas can be used as primary monitoring areas, while the change cycle of the personnel activity area within the area is short, so the corresponding areas can be used as secondary monitoring areas. By conducting regional division, it is convenient for subsequent data fusion in the video stream, and according to the collection of local area video streams, the data in the video stream during transmission can be reduced, thereby improving the transmission efficiency of the video stream and thus improving the monitoring efficiency;

[0054] The video storage module registers the monitoring end corresponding to the monitoring point, establishes a communication channel between the monitoring end and the management end, and stores each area video stream to the management end based on the communication channel;

[0055] Register a monitoring terminal corresponding to the monitoring point, establish a communication channel between the monitoring terminal and the management terminal, and the steps of storing video streams of each area in the management terminal based on the communication channel include: setting multiple sub-databases in the management terminal, and establishing a mapping relationship between the sub-databases and the monitoring areas; determining the target video stream to be transmitted, and determining the sub-database corresponding to the target video stream according to the mapping relationship; establishing a communication channel between the monitoring terminal and the sub-database, and obtaining the transmission frequency corresponding to the communication channel, where the transmission frequency includes the initial transmission frequency or the target transmission frequency, and transmitting the target video stream to the corresponding sub-database based on the transmission frequency, where the target video stream is the first target video stream or the second target video stream; extracting the initial network points corresponding to the active network points of the second target video stream in the first target video stream, and replacing the initial network points in the first target video stream with the active network points of the second target video stream to obtain the area target video stream at the latest time node, storing each area target video stream in the corresponding sub-database, and performing timestamp synchronization encoding on multiple area video streams at the same time node;

[0056] Specifically, set multiple sub-databases in the management terminal, each sub-database corresponds to a monitoring area, and the sub-database is bound to the monitoring area; determine the target video stream to be transmitted in the monitoring area, store the target video stream in the corresponding sub-database, and extract the spatio-temporally aligned video streams from multiple sub-databases according to the mapping relationship, that is, determine the sub-database corresponding to the monitoring area according to the mapping relationship, and then extract the target video streams at the same time node, and perform fusion to complete the restoration of the video to obtain the video situation of each monitoring area before transmission. The fusion process includes the fusion of video streams between each monitoring area and the fusion between the second target video stream and the first target video stream, so that the video streams of each monitoring area can be fused to construct a three-dimensional model, and the video showing the situation of the park can be obtained. The video stream corresponding to the monitoring area with changes can be transmitted, and only the video stream with changes is transmitted, reducing the transmission volume of the park video stream. At the same time, the video streams of each monitoring area are transmitted separately, improving the storage efficiency of the park video during storage;

[0057] The steps of determining the target video stream to be transmitted include: determining the target monitoring area, marking the actual feature points in the target monitoring area, marking virtual feature points corresponding to the actual feature points in the video stream, and connecting the multiple virtual feature points to form a virtual feature network; obtaining the area video stream corresponding to the initial virtual feature network as the first target video stream to be transmitted; obtaining the initial virtual feature network by obtaining the corresponding virtual feature network in the first target video stream; identifying the active network points from the initial virtual feature network to obtain the target grid block, and using the local video stream corresponding to the target grid block as the second target video stream to be transmitted;

[0058] Specifically, there is a physically aligned relationship between the virtual feature network and the actual feature network. By dividing the monitoring area into grid points and marking the corresponding actual feature points, an actual feature network is formed based on the actual feature points. Then, virtual feature points are marked in the video stream corresponding to the actual feature network to generate a virtual feature network, enabling the change situation of the corresponding actual feature points to be understood at any time. The actual feature points can be marked according to physical facilities, personnel conditions, etc. within the monitoring area, and the target grid block composed of active grid points is extracted. An active grid point refers to a grid point where the information corresponding to the feature point changes. For example, if a feature point marks a certain person in the park, when the location information or behavior status information of this person changes, the grid point corresponding to this person is determined as an active grid point. If a feature point marks a physical facility in the park, then when the location or placement status, etc. information of this physical facility changes, the grid point corresponding to this physical facility is determined as an active grid point. It is possible to first transmit the overall video stream as the initial video stream corresponding to the monitoring area. In the subsequent transmission of the video stream, only the video stream corresponding to the feature points with changes is transmitted, and the position where the feature point is located in the initial video stream is replaced, thus completing the transmission of the overall target video stream. For the feature points without changes, there is no need for multiple transmissions, which can reduce the transmission volume of the video stream, thereby improving the transmission efficiency of uploading the video stream to the monitoring management end;

[0059] Obtain the transmission frequency corresponding to the communication channel. Among them, the transmission frequency includes the initial transmission frequency or the target transmission frequency. The step of transmitting the target video stream to the corresponding sub-database based on the transmission frequency includes: Here, the transmission refers to the case where the target transmission frequency is used as the transmission frequency, that is, the transmission step of the video stream in different time periods of each monitoring area with abnormal events after risk assessment. The specific transmission content is to obtain the target transmission frequency corresponding to each communication channel, and based on the target transmission frequency, the abnormal risk index of each monitoring area is determined. Based on the abnormal risk index, the abnormal monitoring areas involved in each abnormal event are sorted by priority. The higher the abnormal risk index of a monitoring area, the higher its corresponding priority. Then, the video stream corresponding to the monitoring area with a large abnormal risk index is preferentially transmitted, so as to achieve hierarchical transmission of the video streams corresponding to multiple abnormal events when multiple abnormal events occur simultaneously, and then achieve off-peak transmission. For example, at the same time node, there are many active grid points in many areas that need to be transmitted using the corresponding communication channels, which will result in a transmission congestion situation. At this time, according to the risk levels of different areas, different preferential transmissions are carried out to improve the transmission efficiency of the video stream;

[0060] The monitoring and early warning module constructs a three-dimensional model of the park based on multiple regional video streams, conducts risk assessment on the regional video streams based on the three-dimensional model of the park to obtain a risk assessment result, and obtains the risk level and event scope of the monitoring area corresponding to the risk assessment result exceeding the preset conditions for early warning;

[0061] The monitoring and early warning module includes:

[0062] A model construction unit, configured to extract spatio-temporally aligned regional video streams from multiple sub-databases and construct a three-dimensional model of the park based on the regional video streams;

[0063] The steps of constructing a three-dimensional model of the park based on the regional video streams include: extracting spatio-temporally aligned regional video streams from multiple sub-databases respectively, marking the regional boundaries of the monitoring regions corresponding to the multiple regional video streams, and assigning a unique identifier to the regional boundary of each monitoring region to obtain a boundary identifier; fusing the multiple regional video streams according to the unique identifier to generate a park video stream; extracting the three-dimensional point cloud in the park video stream, associating the boundary identifier in the regional video stream with the three-dimensional point cloud to generate a primary three-dimensional model of the park, extracting texture information from the video stream, and fitting it to the surface of the primary three-dimensional model of the park to obtain a three-dimensional model of the park;

[0064] Specifically, match the overlapping or adjacent boundaries in multiple video streams based on the boundary ID, and select the priority boundary according to the camera resolution, viewing angle range or model confidence. Perform weighted fusion on the boundary coordinates of the overlapping region (weight = confidence). Superimpose the fused boundary information on the original video stream to generate a park video stream, extract the three-dimensional point cloud from multiple perspective video frames, associate the boundary ID in the video stream with the three-dimensional point cloud to generate a three-dimensional model with boundary labels, extract texture information from the video frames, and fit it to the surface of the three-dimensional model to obtain a three-dimensional model of the park with boundary markings, fuse multiple camera video streams to generate a three-dimensional model of the park, assist in the security patrol and facility management of the park. For example, deploy 10 monitoring cameras to cover the key areas of the park. Identify the boundaries of roads and buildings and assign unique IDs. Fuse multiple video streams according to the IDs to generate a park video stream. Extract the point cloud based on COLMAP to construct a three-dimensional model with boundary labels. Integrate the model into the park management platform to support real-time viewing and interaction, and ensure the time consistency of multiple video streams through timestamp matching;

[0065] Construct a three-dimensional model of the park based on multiple regional video streams, and perform risk assessment on the regional video streams based on this model, and give early warnings on the risk levels and event ranges of the monitoring regions corresponding to the risk assessment results exceeding the preset conditions. By extracting spatio-temporally aligned regional video streams from multiple sub-databases to construct a three-dimensional model, the situation of the park can be presented more intuitively, and at the same time, abnormal event information can be accurately extracted and the risk index can be evaluated, realizing active early warning, discovering potential safety hazards in advance, and effectively reducing the accident rate;

[0066] A risk assessment unit, configured to extract abnormal event information in the three-dimensional model of the park and evaluate the risk index of the corresponding abnormal event according to the abnormal event information;

[0067] Specifically, obtain the area information corresponding to the monitored area from the regional video stream, and determine the abnormal area and the corresponding abnormal information according to the area information. The abnormal information includes the location where the abnormal event occurs, the time when the abnormal event occurs, the duration of the abnormal event, and the frequency of the abnormal event. The formula for evaluating the risk index of each monitored area according to the abnormal information of each monitored area is RI = w1 × location weight + w2 × time weight + w3 × duration weight + w4 × frequency weight. The location where the abnormal event occurs refers to the geographical coordinates or area identifier (such as the camera number). The time when the abnormal event occurs refers to the timestamp accurate to seconds. The duration of the abnormal event refers to the time interval from the detection of the abnormality to the current time. The frequency of the abnormal event refers to the number of times the abnormal event is triggered per unit time. RI represents the risk index (Risk Index), which is a quantitative indicator for comprehensively evaluating the abnormal risk of the monitored area. The higher the value, the greater the risk. w1 represents the location weight coefficient, which is used to reflect the contribution of the location where the abnormality occurs to the risk (for example, the weight of high-risk areas is higher). w2 represents the time weight coefficient, which is used to reflect the contribution of the time when the abnormal event occurs to the risk (for example, the abnormal risk at night is higher). w3 represents the duration weight coefficient, which is used to reflect the contribution of the duration of the abnormality to the risk (the longer the duration, the higher the risk); w4 represents the frequency weight coefficient, which is used to reflect the contribution of the frequency of the abnormal event to the risk (the higher the frequency of the abnormality, the higher the risk). Flexibly adjust the contribution of each factor to the risk to achieve dynamic and accurate risk assessment;

[0068] The abnormal warning unit is used to obtain the area information of the abnormal event corresponding to the risk index exceeding the preset index threshold according to the risk index of the abnormal event. The area information includes the abnormal event range and the abnormal risk index; use the monitored area corresponding to the abnormal event range as the abnormal monitoring area and issue a warning;

[0069] The frequency adjustment module obtains the sub-database with a mapping relationship with the abnormal monitoring area, obtains the communication channel between the abnormal monitoring area and the sub-database, and adjusts the transmission frequency of the communication channel between the abnormal monitoring area and the sub-database to obtain the target transmission frequency;

[0070] The steps of adjusting the transmission frequency of the communication channel between the abnormal monitoring area and the sub-database include: obtaining the defined channel of the communication channel between the abnormal monitoring area and the sub-database, and determining the target transmission frequency corresponding to each defined channel; obtaining multiple independent communication channels, and extracting multiple defined channels triggered by the same abnormal event from the multiple communication channels to obtain a channel group, and activating all the defined channels in the channel group at the same time to start the transmission of the target video stream; setting the transmission frequencies of all the defined channels in the channel group to the same target transmission frequency;

[0071] Specifically, multiple independent communication channels are obtained, each of which uniquely corresponds to a monitoring area, forming a "channel-area" mapping relationship. Basic transmission parameters (such as default frequency, bandwidth, and encoding format) are configured for each channel to ensure that the channel has the ability to operate independently. Abnormal events (such as intrusion, equipment failure, and natural disasters) and the set of monitoring areas affected by them are captured in real time. According to the impact range of the abnormal event, the corresponding communication channels are screened to form an "event-channel" association list. For multiple communication channels caused by the same abnormal event, the following operations are performed: Synchronous start: Activate all limited channels in the channel group at the same time to start video stream transmission; Frequency unification: Set the transmission frequency in the channel group to the same target value (such as dynamically calculated based on bandwidth restrictions or event priorities); Association binding: Dynamically establish an association relationship within the channel group to form a logical communication group to ensure that the frame-level synchronization and timing of all video streams in the channel group are consistent. Real-time monitoring of the channel group status ensures that the synchronous transmission is continuously effective; when the abnormal event is processed or the trigger condition fails, the channel association is automatically released and the independent transmission mode is restored. Based on the impact range of the abnormal event, the relevant communication channels are dynamically screened and bound; Supports concurrent processing of multiple events, and the channel groups corresponding to different events are independent of each other. Conditions for disassociating: abnormal events have been handled; with abnormal events as the core, relevant communication channels are automatically screened and associated to achieve rapid response and collaborative work.

[0072] Specifically, during the adjustment process, the adjustment can be made according to the number of limited channels. When it is determined that the number of limited channels is 1, the transmission frequency of the limited channel is directly adjusted to the target transmission frequency corresponding to the limited channel; if it is determined that the number of limited channels is not 1, all the limited channels are extracted, and the target transmission frequency is used as the transmission frequency of all the limited channels according to the association relationship between the limited channels; the number of limited channels can be determined according to how many monitoring areas are involved in the scope of the abnormal event. When multiple monitoring areas are involved, it means that there are multiple abnormal areas, and multiple abnormal areas correspond to multiple communication channels to obtain multiple limited channels, which means that the number of limited channels is not 1. When the scope of the abnormal event is only located in one monitoring area, it means that only one limited channel is needed to complete the transmission of the corresponding video stream, and then only the transmission frequency of the limited channel is adjusted. The degree to which the transmission frequency is adjusted can be determined according to the abnormal risk level of the abnormal event. The higher the abnormal risk level, the higher the transmission frequency, and the lower the abnormal risk level, the lower the transmission frequency. The lowest transmission frequency needs to be greater than or equal to the initial transmission frequency.

[0073] The step of determining the target transmission frequency corresponding to each limited channel includes: classifying the abnormal risk index into multiple risk levels, and setting the transmission frequency corresponding to the risk level; obtaining the abnormal risk index of the abnormal monitoring area, and obtaining the transmission frequency corresponding to the risk level to which the abnormal risk index belongs as the target transmission frequency of the corresponding limited channel;

[0074] Specifically, the abnormal risk index of an abnormal event applies to the abnormal risk indices of multiple monitoring areas involved in the scope of the abnormal event corresponding to the abnormal event. The abnormal risk index evaluated from an abnormal event can be used to select the target transmission frequency corresponding to multiple abnormal monitoring areas involved in the scope of the abnormal event. For example, if an abnormal event involves two abnormal monitoring areas, it means that the defined channels corresponding to these two abnormal monitoring areas need to be adjusted to the same target transmission frequency. Since the target transmission frequency is determined by the abnormal risk index, and the abnormal risk index is evaluated based on the regional information of the abnormal event;

[0075] In the present invention, by obtaining the park area information, including the area function information and the area change cycle, the park is divided into multiple monitoring areas based on the area function information, and then each monitoring area is further divided into multiple sub-monitoring areas based on the area change cycle. Then, monitoring points are reasonably arranged in each monitoring area. The refined division and arrangement methods can ensure that the monitoring points can more effectively cover key areas, reduce monitoring blind spots, and improve the comprehensiveness and pertinence of monitoring. Multiple sub-databases are set at the management end, and a mapping relationship is established between the sub-databases and the monitoring areas. By determining the target video stream to be transmitted, it is transmitted to the corresponding sub-database according to the mapping relationship, and timestamp synchronization coding is performed on the video streams of multiple areas at the same time node. It can make the management of video data more orderly, facilitate the quick and accurate retrieval and call of video data in a specific area, and improve the monitoring efficiency and management level. Obtain the sub-database that has a mapping relationship with the abnormal monitoring area, and obtain the communication channel between the abnormal monitoring area and the sub-database. Adjust the transmission frequency of the communication channel between the abnormal monitoring area and the sub-database to obtain the target transmission frequency, and set the transmission frequencies of all the defined channels in the channel group to the same target transmission frequency. The way of dynamically adjusting the communication frequency can ensure that the video transmission in important areas is guaranteed with sufficient communication resources, and at the same time avoid unnecessary waste of communication resources, and improve the utilization efficiency of communication resources.

[0076] In the description of this specification, the descriptions referring to the terms "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0077] Although embodiments of the present invention have been shown and described, those of ordinary skill in the art will appreciate that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A remote monitoring system for a park based on three-dimensional digitization, characterized in that Including: A video acquisition module that divides the monitored areas in the park into multiple monitored areas and arranges monitoring points in each monitored area; Collecting video streams of each monitored area through the monitoring points to obtain multiple area video streams; A video storage module that registers a monitoring end corresponding to the monitoring point, establishes a communication channel between the monitoring end and the management end, and stores each area video stream to the management end based on the communication channel; A monitoring and warning module that constructs a three-dimensional model of the park based on multiple area video streams, performs risk assessment on the area video streams based on the three-dimensional model of the park to obtain a risk assessment result, and obtains the risk level and event scope of the monitored area corresponding to the risk assessment result exceeding the preset conditions for warning; A frequency adjustment module that obtains a sub-database having a mapping relationship with the abnormal monitored area, obtains a communication channel between the abnormal monitored area and the sub-database, and adjusts the transmission frequency of the communication channel between the abnormal monitored area and the sub-database to obtain a target transmission frequency.

2. The park remote monitoring system based on three-dimensional digitization according to claim 1, wherein: The step of dividing the monitored areas in the park into multiple monitored areas and arranging monitoring points in each monitored area includes: Obtaining the area information of the park to be monitored, where the area information of the park includes area function information and area change cycle; Dividing the park into multiple monitored areas based on the area function information, and performing secondary area division on each monitored area based on the area change cycle to obtain multiple sub-monitored areas, where the sub-monitored areas include primary monitored areas and secondary monitored areas; Arranging monitoring points in each monitored area.

3. The park remote monitoring system based on three-dimensional digitization according to claim 1, wherein: The step of registering a monitoring end corresponding to the monitoring point, establishing a communication channel between the monitoring end and the management end, and storing each area video stream to the management end based on the communication channel includes: Setting multiple sub-databases in the management end and establishing a mapping relationship between the sub-databases and the monitored areas; Determining the target video stream to be transmitted, and determining the sub-database corresponding to the target video stream according to the mapping relationship; Establishing a communication channel between the monitoring end and the sub-database, and obtaining the transmission frequency corresponding to the communication channel, where the transmission frequency includes an initial transmission frequency or a target transmission frequency; Transmitting the target video stream to the corresponding sub-database based on the transmission frequency, where the target video stream is a first target video stream or a second target video stream; Extracting the initial network points corresponding to the active network points of the second target video stream in the first target video stream, and replacing the initial network points in the first target video stream with the active network points of the second target video stream to obtain the area target video stream at the latest time node; Storing each area target video stream in the corresponding sub-database, and performing timestamp synchronization coding on multiple area video streams at the same time node.

4. The park remote monitoring system based on three-dimensional digitization according to claim 3, characterized in that: The step of determining the target video stream to be transmitted includes: Determining the target monitored area, marking the actual feature points in the target monitored area, marking virtual feature points corresponding to the actual feature points in the video stream, and connecting the multiple virtual feature points to form a virtual feature network; Obtaining the area video stream corresponding to the initial virtual feature network as the first target video stream to be transmitted; Obtaining the initial virtual feature network by obtaining the corresponding virtual feature network in the first target video stream; Identify the active grid points from the initial virtual feature network to obtain the target grid block, and use the local video stream corresponding to the target grid block as the second target video stream to be transmitted.

5. The park remote monitoring system based on three-dimensional digitization according to claim 1, wherein: The monitoring and early warning module includes: A model construction unit, configured to extract spatio-temporally aligned regional video streams from multiple sub-databases and construct a 3D model of the park based on the regional video streams; A risk assessment unit, configured to extract abnormal event information from the 3D model of the park and evaluate the risk index of the corresponding abnormal event according to the abnormal event information; An abnormal early warning unit, configured to obtain the regional information of the abnormal event corresponding to the risk index exceeding the preset index threshold according to the risk index of the abnormal event, where the regional information includes the abnormal event range and the abnormal risk index; use the monitoring area corresponding to the abnormal event range as the abnormal monitoring area and give an early warning.

6. The park remote monitoring system based on three-dimensional digitization according to claim 5, characterized in that: The steps of constructing a 3D model of the park based on the regional video streams include: Extract spatio-temporally aligned regional video streams from multiple sub-databases respectively, mark the regional boundaries of the monitoring areas corresponding to the multiple regional video streams, and assign a unique identifier to the regional boundary of each monitoring area to obtain a boundary identifier; Fuse the multiple regional video streams according to the unique identifier to generate a park video stream; Extract the 3D point cloud from the park video stream, associate the boundary identifier in the regional video stream with the 3D point cloud, and generate a primary 3D model of the park; Extract texture information from the video stream and fit it to the surface of the primary 3D model of the park to obtain the 3D model of the park.

7. The park remote monitoring system based on three-dimensional digitization according to claim 1, wherein: The steps of adjusting the transmission frequency of the communication channel between the abnormal monitoring area and the sub-database include: Obtain the communication channel between the abnormal monitoring area and the sub-database to obtain a defined channel, and determine the target transmission frequency corresponding to each defined channel; Obtain multiple independent communication channels, and extract multiple defined channels triggered by the same abnormal event from the multiple communication channels to obtain a channel group, activate all the defined channels in the channel group at the same time, and start the transmission of the target video stream; Set the transmission frequencies of all the defined channels in the channel group to the same target transmission frequency.

8. The three-dimensional digitalized park remote monitoring system according to claim 1 is characterized in that: The steps of determining the target transmission frequency corresponding to each defined channel include: Divide the abnormal risk index into multiple risk levels, and set the transmission frequency corresponding to the risk level; Obtain the abnormal risk index of the abnormal monitoring area, and obtain the transmission frequency corresponding to the risk level to which the abnormal risk index belongs as the target transmission frequency of the corresponding defined channel.

Images