A method for intelligent early warning and disposal of high-position videos at the city level

Abstract

The present invention discloses a method for intelligent early warning and disposal of high-position videos at the city level, including: the construction process of parking space perception devices, the process of reporting parking space status, the intelligent early warning process of the management background, the process of processing early warning work orders, and the process of counting high-position video recognition failures. The method for intelligent early warning and disposal of high-position videos at the city level provided by the present invention conducts precise intelligent early warning on the failure situations of parking records uploaded by high-position videos, so that manual processing can be carried out in a timely manner. And through the statistical analysis of the error rate based on the fault marks after processing, the recognition rate of high-position video devices and the algorithm of software intelligent early warning are continuously optimized and improved, ensuring the arrears records caused by the device accuracy for vehicle owners and reducing the patrol feedback of front-end toll collectors to reduce labor costs.

Description

Technical Field

[0001] The present invention relates to the technical field of high-position video monitoring, and in particular to a method for intelligent early warning and disposal of city-level high-position video. Background Art

[0002] Through information technology, front-end sensing equipment (high-position video) is built to obtain real-time information on road parking spaces, detect the occupancy status of parking spaces, and transmit the parking space occupancy status information (parking start time (accurate to seconds), end time (accurate to seconds), parking space number, license plate number, etc.) to the background management system in real time and reliably through the network. Since there are errors in the recognition of the equipment, the uploaded data is not accurate enough, so it is particularly important for the background system to analyze and process these data accordingly to realize intelligent early warning and disposal methods. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a method for city-level high-position video intelligent early warning and disposal, which solves the errors existing in the parking space sensing device when the parking space information is installed by the device by analyzing the data transmitted by the device, and corrects the error data through the intelligent early warning processing method to make the parking space information more accurate.

[0004] The present invention is achieved in that:

[0005] A city-level high-position video intelligent early warning and disposal method, including: parking space sensing equipment construction process, parking space status reporting process, management background intelligent early warning process, early warning work order processing process and high-position video statistical fault identification process;

[0006] The process of building the parking space sensing equipment includes: setting up poles and installing intelligent gun cameras on the same side or opposite side of multiple parking spaces in the same area to detect and capture vehicles entering and leaving the parking spaces in real time;

[0007] The parking space status reporting process includes: acquiring parking space information through high-position video, and then transmitting it to the management backend in real time through the network;

[0008] The management backend intelligent early warning process includes: intelligently generating an early warning work order according to preset early warning conditions;

[0009] The warning work order processing process includes: analyzing the warning work order type and intelligently correcting the parking record according to the result of manually processing the warning work order, including supplementing information, invalidating or closing the work order, and modifying the parking space information;

[0010] The process of counting high-position video recognition faults includes: when obtaining the early warning work order for manual processing, identifying the fault type of high-position video recognition errors, managing the fault statistics in the background, calculating the high-position video failure rate, and analyzing the cause of the failure.

[0011] Further, during the reporting process of the parking space status, the parking berth information includes: parking start time, parking end time, berth number, license plate number, and pictures during the parking process.

[0012] Further, the intelligent early warning process of the management background specifically includes:

[0013] When there is only a record of a vehicle leaving a parking berth, generate an early warning work order of the type "entry not reported";

[0014] When a second vehicle enters a parking berth and the previous record of this parking berth is the entry record of the first vehicle, generate an early warning work order of the type "exit not reported or multiple vehicles in one berth", and set the parking status of the first vehicle as abnormal or invalid;

[0015] When a vehicle enters the first berth without an exit record and then enters the second berth, generate an early warning work order of the type "one vehicle in multiple berths".

[0016] Further, the processing process of the early warning work order specifically includes:

[0017] When processing the "entry not reported" early warning work order, if the vehicle has an actual entry, perform the "supplement information" process to supplement the vehicle's entry time and entry pictures; if the vehicle has no actual entry, perform the "set as invalid" process to close the "entry not reported" early warning work order;

[0018] When processing the "exit not reported or multiple vehicles in one berth" early warning work order, if the result of manual processing is "exit not reported", perform the "supplement information" process, and automatically match the exit time of the first vehicle with the entry time of the second vehicle;

[0019] When processing the "exit not reported or multiple vehicles in one berth" early warning work order, if the result of manual processing is "multiple vehicles in one berth" and the second vehicle has no actual entry, perform the "set as invalid" process to set the second vehicle as invalid, and then close the work order; if the result of manual processing is "multiple vehicles in one berth" and the second vehicle is actually parked in the second berth, perform the "modify berth information" process to change the parking space of the second vehicle to the second berth;

[0020] When processing the "one vehicle in multiple berths" early warning work order, if the vehicle has a payment record in one of the berths, automatically set the other berth as abnormal or invalid, so that when the vehicle enters other berths again, it will no longer report the "one vehicle in multiple berths" work order;

[0021] If the vehicle has no payment record and the result of manual processing is that the first vehicle enters from the first berth, fails to report when leaving, and then enters from the second berth, "supplementary information" processing is performed to supplement the entry time and entry picture of the first vehicle.

[0022] If the vehicle has no payment record and the result of manual processing is that the first vehicle actually enters from the first berth / second berth, the parking information of the first vehicle at the second berth / first berth is processed as "set to invalid".

[0023] Furthermore, during the process of counting high-position video recognition failures, the failure types include failure to report entry, failure to report exit, incorrect parking space recognition, and misaligned capture.

[0024] The statistical data includes the number of vehicle entries, the number of vehicle exits, the number of unreported entries, the number of unreported exits, the number of incorrect parking space recognitions, the number of misaligned captures, and the number of recognition failures; among them, the number of recognition failures includes the number of unreported entries, the number of unreported exits, the number of incorrect parking space recognitions, and the number of misaligned captures.

[0025] The high-position recognition failure rate = the number of recognition failures / (the number of vehicle entries + the number of vehicle exits + the number of recognition failures).

[0026] The present invention has the following advantages:

[0027] Through precise intelligent early warning of the failure situations of the parking records uploaded by the high-position video, manual processing is carried out in a timely manner, and through the failure marks after processing, statistical analysis of the error rate is carried out, continuously optimizing and improving the recognition rate of the high-position video equipment and the algorithm of the software intelligent early warning, ensuring the arrears records caused by the equipment accuracy for the vehicle owners and reducing the patrol feedback of the front-end toll collectors to reduce the labor cost. Brief Description of the Drawings

[0028] The following further describes the present invention with reference to the accompanying drawings in conjunction with embodiments.

[0029] Figure 1 is the execution flowchart of the method of the embodiment of the present invention;

[0030] Figure 2 is the schematic diagram of the early warning types in the embodiment of the present invention. Detailed Embodiments

[0031] The present invention provides a method for intelligent early warning and disposal of urban-level high-position video. By analyzing the data transmitted by the equipment for intelligent early warning and disposal, the error existing in the equipment parking space information is solved, and the error data is corrected through the method of intelligent early warning processing to make the parking space information more accurate.

[0032] As Figure 1 shown, the overall implementation process of the present invention is as follows:

[0033] S1. Construction of parking space sensing equipment: Set up poles and install intelligent gun cameras on the same side or opposite side of the parking spaces on urban roads to effectively cover the parking space area as much as possible, and conduct real-time detection and capture of vehicles parking in and out.

[0034] S2. Parking space status reporting: The high-position video captures parking space parking record information (parking start time (accurate to seconds), parking end time (accurate to seconds), parking space number, license plate number, parking process pictures, etc.) and transmits it to the background management system in real time and reliably through the network.

[0035] S3. Intelligent warning in the management background: The management background analyzes the parking records uploaded by the high-position video, determines the type of fault in the parking space recognition of the smart gun camera, and makes a warning.

[0036] First, we analyze the scenarios where high-position video recognition failures occur, build early warning models based on the scenarios, and make corresponding early warning work orders based on different scenarios. The types of early warning work orders include: unreported entry, unreported exit, multiple vehicles for one person, and multiple positions for one vehicle.

[0037] like Figure 2 As shown, the parking spaces are 4 parallel parking spaces. The warning types and handling methods can be as follows:

[0038]

[0039]

[0040] S4. Processing warning work orders: Intelligent warning generates work orders, the management background analyzes the types of warning work orders, intelligently processes work orders in some scenarios and provides data for manual processing of work orders, to ensure the accuracy of work order warnings and improve the efficiency of manual processing.

[0041] The intelligent processing method is as follows:

[0042] 1. Unreported appearance or multiple vehicles per vehicle

[0043] Car A's exit from parking space 001 was not reported. After the first subsequent car (car B) entered, a work order "exit was not reported or there were multiple cars in one parking space" was reported. Car B exited normally, and car A was set to be abnormal or invalid. After the second car (car C) entered the parking space, no work order was reported.

[0044] 2. One car with multiple seats

[0045] If the vehicle has a payment record in one of the parking spaces, the other parking space will be automatically set as abnormal or invalid. When the vehicle enters the other parking space again, the "one vehicle in multiple spaces" work order will no longer be reported.

[0046] 3. Data on unreported entry and unreported exit

[0047] a. When the type of the warning work order is "not reported upon entry", the parking record only has the departure time but no entry time of the vehicle. The management platform automatically matches the departure time of the previous vehicle on the work order, and during manual processing, the actual entry time of the vehicle can be quickly located and corrected.

[0048] b. When the type of the warning work order is "not reported upon departure", the parking record only has the entry time but no departure time of the vehicle. The management platform automatically matches the entry time of the next vehicle on the work order, and during manual processing, the actual departure time of the vehicle can be quickly located and corrected.

[0049] S5. Statistic of high-position video recognition failures: When processing the work order manually, mark the failure type for the high-position video recognition error. The management background statistically analyzes the failures, calculates the high-position video recognition failure rate, and analyzes the failure causes to provide data support for the optimization of the high-position video recognition rate and the intelligent warning algorithm.

[0050] The failure types include: not reported upon entry, not reported upon departure, wrong parking space recognition, misaligned capture

[0051] The statistical data includes: the number of vehicle entries, the number of vehicle departures, the number of unreported entries, the number of unreported departures, the number of wrong parking space recognitions, the number of misaligned captures, the number of recognition failures; among them, the number of recognition failures includes (the number of unreported entries, the number of unreported departures, the number of wrong parking space recognitions, the number of misaligned captures)

[0052] The high-position recognition failure rate = the number of recognition failures / (the number of vehicle entries + the number of vehicle departures + the number of recognition failures)

[0053] As Figure 1 shown, a specific embodiment of the present invention is as follows:

[0054] A method for intelligent warning and disposal of urban high-position video includes: the process of building parking space perception devices, the process of reporting parking space status, the process of intelligent warning in the management background, the process of processing warning work orders, and the process of statistically analyzing high-position video recognition failures;

[0055] The process of building parking space perception devices includes: installing intelligent cameras on the same side or opposite sides of multiple parking berths in the same area, and performing real-time detection and capture of vehicles entering and leaving the parking berths;

[0056] The process of reporting parking space status includes: obtaining the parking berth information collected by high-position video (such as the start time of parking, the end time of parking, the berth number, the license plate number, and the pictures during the parking process), and then transmitting it to the management background in real time through the network;

[0057] The process of intelligent warning in the management background includes: intelligently generating warning work orders according to the preset warning conditions;

[0058] The process of processing the warning work order includes: analyzing the type of the warning work order and intelligently correcting the parking record according to the result of manually processing the warning work order, including supplementing information, setting it as invalid or closing the work order, and modifying the berth information;

[0059] The process of statistically analyzing the high-position video recognition failure includes: when manually processing the warning work order, obtaining the failure types identified by the high-position video recognition error, and the management background statistically analyzes the failure, calculates the high-position video failure rate, and analyzes the cause of the failure.

[0060] Specifically, the intelligent warning process of the management background includes:

[0061] When there is only a record of a vehicle leaving a parking berth, a warning work order of the type "entry not reported" is generated;

[0062] When a second vehicle enters a parking berth and the previous record of this berth is the entry record of the first vehicle, a warning work order of the type "exit not reported or multiple vehicles in one berth" is generated, and the parking status of the first vehicle is set as abnormal or invalid;

[0063] When a vehicle enters a first berth (the first berth is any berth in this area) without an exit record and then enters a second berth (the second berth is any berth in this area except the first berth), a warning work order of the type "one vehicle in multiple berths" is generated.

[0064] The process of processing the warning work order specifically includes:

[0065] When processing the warning work order of "entry not reported", if the vehicle actually enters, the "supplement information" process is executed to supplement the entry time and entry picture of the vehicle; if the vehicle does not actually enter, the "set as invalid" process is executed to close the warning work order of "entry not reported";

[0066] When processing the warning work order of "exit not reported or multiple vehicles in one berth", if the result of manual processing is "exit not reported", the "supplement information" process is executed, and the exit time of the first vehicle is automatically matched with the entry time of the second vehicle;

[0067] When processing the warning work order of "exit not reported or multiple vehicles in one berth", if the result of manual processing is "multiple vehicles in one berth" and the second vehicle does not actually enter, the "set as invalid" process is executed to set the second vehicle as invalid, and then the work order is closed; if the result of manual processing is "multiple vehicles in one berth" and the second vehicle actually parks in the second berth, the "modify berth information" process is executed to change the parking space of the second vehicle to the second berth;

[0068] When processing the warning work order of "one vehicle in multiple berths", if the vehicle has a payment record in one of the berths, the other berth will be automatically set as abnormal or invalid, so that when the vehicle enters other berths again, the "one vehicle in multiple berths" work order will not be reported;

[0069] If the vehicle has no payment record and the result of manual processing is that the first vehicle enters from the first berth, fails to report when leaving, and then enters from the second berth, the "supplementary information" process will be executed to supplement the entry time and entry picture of the first vehicle;

[0070] If the vehicle has no payment record and the result of manual processing is that the first vehicle actually enters from the first berth / second berth, the parking information of the first vehicle in the second berth / first berth will be processed as "set to invalid".

[0071] In a possible implementation manner, during the process of counting high-position video recognition failures, the failure types include failure to report entry, failure to report exit, parking space recognition error, and misaligned capture;

[0072] The statistical data includes the number of parking entries, the number of parking exits, the number of unreported entries, the number of unreported exits, the number of parking space recognition errors, the number of misaligned captures, and the number of recognition failures; among them, the number of recognition failures includes the number of unreported entries, the number of unreported exits, the number of parking space recognition errors, and the number of misaligned captures;

[0073] The high-position recognition failure rate = the number of recognition failures / (the number of vehicle entries + the number of vehicle exits + the number of recognition failures).

[0074] The present invention performs precise intelligent warning on the failure situations of the parking records uploaded by the high-position video, so that manual processing can be carried out in a timely manner, and through the statistical analysis of the error rate with the processed failure marks, the recognition rate of the high-position video device and the algorithm of the software intelligent warning are continuously optimized and improved, ensuring the arrears records caused by the equipment accuracy for the vehicle owners and reducing the inspection feedback of the front-end toll collectors to reduce the labor cost.

[0075] Although the specific implementation manners of the present invention are described above, those skilled in the art of this technology should understand that the specific embodiments we described are illustrative rather than used to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be covered by the scope protected by the claims of the present invention.

Claims

1. A method for intelligent early warning and disposal of high-position videos at the city level, characterized in that, Including: The construction process of parking space perception equipment, the process of reporting parking space status, the intelligent early warning process of the management background, the process of processing early warning work orders, and the process of statistically high-position video recognition failures; The construction process of the parking space perception equipment includes: installing intelligent cameras on the same side or opposite side of multiple parking berths in the same area, and performing real-time detection and capture of vehicles entering and leaving the parking berths; The process of reporting parking space status includes: obtaining high-position video to collect parking berth information, and then transmitting it to the management background in real time through the network; The intelligent early warning process of the management background includes: intelligently generating early warning work orders according to preset early warning conditions; The process of processing early warning work orders includes: analyzing the type of early warning work orders and intelligently correcting parking records according to the results of manually processing the early warning work orders, including supplementing information, setting as invalid or closing the work order, and modifying berth information; The process of statistically high-position video recognition failures includes: obtaining the fault types identified by high-position video recognition when manually processing the early warning work orders, the management background statistically analyzing the faults, calculating the high-position video failure rate, and analyzing the causes of the faults; The intelligent early warning process of the management background specifically includes: When there is only a record of a vehicle leaving a parking berth, generating an early warning work order of the type "entry not reported"; When a second vehicle enters a parking berth and the previous record of this parking berth is the entry record of the first vehicle, generating an early warning work order of the type "exit not reported or multiple vehicles in one berth", and setting the parking status of the first vehicle as abnormal or invalid; When a vehicle enters the first berth without an exit record and then enters the second berth, generating an early warning work order of the type "one vehicle in multiple berths"; The process of processing early warning work orders specifically includes: When processing the "entry not reported" early warning work order, if the vehicle has an actual entry, perform the "supplement information" process to supplement the entry time and entry picture of the vehicle; if the vehicle has no actual entry, perform the "set as invalid" process to close the "entry not reported" early warning work order; When processing the "exit not reported or multiple vehicles in one berth" early warning work order, if the result of manual processing is "exit not reported", perform the "supplement information" process to automatically match the exit time of the first vehicle with the entry time of the second vehicle; When processing the "exit not reported or multiple vehicles in one berth" early warning work order, if the result of manual processing is "multiple vehicles in one berth" and the second vehicle has no actual entry, perform the "set as invalid" process to set the second vehicle as invalid, and then close the work order; if the result of manual processing is "multiple vehicles in one berth" and the second vehicle is actually parked in the second berth, perform the "modify berth information" process to change the parking space of the second vehicle to the second berth; When processing the "one vehicle in multiple berths" early warning work order, if the vehicle has a payment record in one of the berths, automatically set the other berth as abnormal or invalid, so that when the vehicle enters other berths again, it will no longer report the "one vehicle in multiple berths" work order; If the vehicle has no payment record and the result of manual processing is that the first vehicle enters from the first berth, the exit is not reported, and then enters from the second berth, perform the "supplement information" process to supplement the entry time and entry picture of the first vehicle; If the vehicle has no payment record and the result of manual processing shows that the first vehicle actually entered from the first parking space / second parking space, then the parking information of the first vehicle in the second parking space / first parking space shall be processed as "set to invalid".

2. The method according to claim 1, characterized in that: During the reporting process of the parking space status, the parking space information includes: parking start time, parking end time, parking space number, license plate number, and pictures during the parking process.

3. The method according to claim 1, characterized in that: During the process of counting high-position video recognition failures, the types of failures include failure to report entry, failure to report exit, incorrect parking space recognition, and misaligned capture. The statistical data includes the number of vehicle entries, the number of vehicle exits, the number of unreported entries, the number of unreported exits, the number of incorrect parking space recognitions, the number of misaligned captures, and the number of recognition failures; among them, the number of recognition failures includes the number of unreported entries, the number of unreported exits, the number of incorrect parking space recognitions, and the number of misaligned captures. High-position recognition failure rate = number of recognition failures / (number of vehicle entries + number of vehicle exits + number of recognition failures).

Images