Billboard content security monitoring method and system for highways

Abstract

The application relates to the technical field of advertising board content safety monitoring, in particular to an advertising board content safety monitoring method and system for highways. In the system, an advertising violation event analysis module controls a cloud recognition platform to receive advertising board images uploaded by a camera, image feature extraction is carried out on the advertising board images through image recognition technology, content compliance analysis is carried out on the advertising board images, a set of violation events corresponding to the advertising board images is identified, the identified violation events are bound with the corresponding advertising board images, and feedback is given to a display end. The application takes into account the influence of the change of the surrounding environment of the advertising board on the photovoltaic power supply state of the camera, solves the camera power supply endurance problem under continuous severe climate environment, realizes dynamic regulation and control of the camera shooting cycle according to the residual power and the future weather change of the surrounding environment of the advertising board, and prolongs the safety monitoring time length of the camera on the advertising board content.

Description

Technical Field

[0001] This invention relates to the field of billboard content security monitoring technology, specifically to a method and system for monitoring billboard content security on highways. Background Technology

[0002] As vital transportation arteries, highways utilize roadside billboards (including high-rise billboards, overpass billboards, and service area billboards) as important carriers for commercial advertising and public service information dissemination, reaching a wide audience and exerting a significant influence. However, the compliance and safety of the billboard content are of paramount importance. The content may involve false advertising, illegal or irregular information, vulgar or harmful information, or damage or detachment of content that could lead to misleading information. Therefore, it is necessary to monitor the content of highway billboards.

[0003] Traditional methods of monitoring billboard content mainly rely on regular manual inspections and passive reporting. This method is inefficient and lacks timeliness. Faced with a large number of dispersed highway billboards, manual inspections are difficult to achieve real-time and comprehensive monitoring.

[0004] With the development of IoT technology, some road sections use fixed photovoltaic-powered cameras around billboards to transmit images back to the monitoring center. However, this method does not take into account the photovoltaic power supply of the cameras, which often leads to insufficient power supply to the cameras in continuous severe weather conditions, making it impossible to monitor and transmit the billboard content in a timely manner. Furthermore, it cannot dynamically adjust the camera shooting cycle based on the remaining power and future weather changes around the billboard to ensure the camera's endurance. Therefore, the existing technology has significant shortcomings. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for monitoring the safety of billboard content on highways, in order to solve the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for security monitoring of billboard content on highways, the method comprising the following steps:

[0007] S1. Obtain the location information of billboards set up on the highway; periodically take pictures of billboards set up on the highway using deployed photovoltaic-powered cameras, wherein the cameras communicate with the cloud recognition platform through a 5G network and upload the location information of the corresponding cameras and the billboard images taken each time;

[0008] S2. The cloud-based recognition platform receives billboard images uploaded by cameras, extracts image features from the billboard images using image recognition technology, performs content compliance analysis on the billboard images, identifies the set of violations corresponding to the billboard images, and binds the identified violations with the corresponding billboard images, feeding them back to the display end.

[0009] S3. Obtain the current shooting environment information and remaining power value of the photovoltaic-powered cameras deployed on the highway; combine the current weather forecast information for each photovoltaic-powered camera deployed on the highway to assess the battery life crisis coefficient corresponding to each photovoltaic-powered camera deployed on the highway; combine the set of violation events corresponding to the images captured by each deployed billboard in historical data to calculate the feedback demand assessment value of each deployed billboard.

[0010] S4. Based on the battery life crisis coefficient of each photovoltaic-powered camera deployed on the highway and the feedback demand assessment value of each billboard deployed, generate the shooting cycle adjustment coefficient of each photovoltaic-powered camera deployed on the highway, and update the shooting cycle of each photovoltaic-powered camera deployed on the highway in real time.

[0011] Furthermore, the shooting time period for each photovoltaic-powered camera deployed in S1 to capture images of billboards set up on the corresponding highway is obtained by querying a database; different cameras correspond to different shooting time periods.

[0012] During the communication process between the camera and the cloud recognition platform via the 5G network, the uploaded data also includes the time of each shot taken by the corresponding camera, and the billboard image captured each time is bound with the corresponding shooting time and the corresponding camera location information.

[0013] Furthermore, S2 includes:

[0014] S21. The cloud recognition platform receives the billboard images uploaded by the camera and assigns a number to each uploaded billboard image; the number corresponding to the i-th uploaded billboard image is denoted as Bi.

[0015] S22. Extract text information from the billboard image corresponding to Bi using OCR technology, summarize the keywords in the extracted text information, and obtain the image semantic feature set of the billboard image corresponding to Bi.

[0016] S23. Obtain the average gray value of each pixel in the billboard image corresponding to Bi as the gray value of the corresponding billboard image, denoted as HBi; obtain the set of billboard images uploaded by the camera to which Bi belongs within the most recent preset time period, denoted as the reference set of Bi; obtain the billboard image in the reference set of Bi whose absolute value of the difference between the corresponding gray value and HBi is less than the preset gray value threshold and is closest to the shooting time of Bi, denoted as the comparison image of Bi.

[0017] S24. Obtain the difference in grayscale values ​​between the corresponding pixels in the same position between the comparison image corresponding to Bi and the billboard image corresponding to Bi. Mark the positions of pixels whose grayscale value difference is greater than a preset value in the billboard image corresponding to Bi. Use the pixel marking results in the billboard image corresponding to Bi as the image feature set of the billboard image corresponding to Bi.

[0018] S25. Compare each keyword in the semantic feature set of the billboard image corresponding to Bi with the preset keyword library corresponding to the semantic violation. The set of keywords belonging to the preset keyword library corresponding to the semantic violation in the comparison results is taken as the semantic violation judgment set. When the semantic violation judgment set is empty, it is determined that there is no semantic violation event in the billboard image corresponding to Bi; otherwise, it is determined that there is a semantic violation event in the billboard image corresponding to Bi.

[0019] S26. Divide the pixels in the image feature set of the billboard image corresponding to Bi into groups with a pixel distance less than a preset pixel distance, and remove groups with fewer than a preset number of pixels in each group. Record the image feature set of the billboard image corresponding to Bi after performing the removal of groups as the image violation judgment set. When the image violation judgment set is empty, it is determined that there is no image violation event in the billboard image corresponding to Bi; otherwise, it is determined that there is an image violation event in the billboard image corresponding to Bi.

[0020] S27. The sum of the judgment results of the visual violation event and the judgment results of the semantic violation event of the billboard image corresponding to Bi is denoted as the violation event set of the billboard image corresponding to Bi.

[0021] This invention identifies the status of violations of highway billboard content from both semantic and visual perspectives. It identifies risks such as false advertising, illegal information, vulgar and harmful information, and misleading content due to damage or detachment. The collection of violations of billboard images not only provides relevant administrators with auxiliary references for rectifying billboard violations, but also provides reference suggestions and data support for the maintenance of highway billboards.

[0022] Furthermore, the shooting environment information in S3 includes the effective visibility of the environment at the time of shooting; the weather forecast information includes the effective visibility prediction value, solar irradiance prediction value, and wind speed corresponding to each time point in the subsequent preset time period based on the current time.

[0023] The calculation formula involved in S3 for evaluating the battery life crisis factor of each photovoltaic-powered camera deployed on the highway is as follows:

[0024] ;

[0025] Among them, EL j This represents the battery life risk assessment value corresponding to the j-th photovoltaic-powered camera deployed on the highway; TG j TG represents the length of the shooting time interval corresponding to the current time weather forecast information for the j-th photovoltaic-powered camera deployed on the highway; j At any given time point within the corresponding time interval, the predicted effective visibility value of the j-th photovoltaic-powered camera deployed on the highway, based on the current time's weather forecast information, is greater than the first visibility threshold; BP j H represents the remaining power value of the j-th photovoltaic-powered camera deployed on the highway at the current time; j TP represents the power consumption of the j-th photovoltaic-powered camera deployed on the highway during each shot at the current time; j This represents the shooting cycle of the j-th photovoltaic-powered camera deployed on the highway at the current time; TB j G represents the length of the time interval in which the predicted solar irradiance value of the j-th photovoltaic-powered camera deployed on the highway is greater than the preset solar irradiance threshold based on the current time weather forecast information; （j，t） TB j The predicted solar irradiance at time point t within the corresponding time interval; β （j，t） TB j The effective visibility prediction value at time point t within the corresponding time interval is the solar irradiance loss coefficient bound in the database preset form; F{} represents the operation of querying the solar power generation speed corresponding to different solar irradiance in the database preset form;

[0026] This indicates that the solar irradiance in the database's preset form is... The corresponding speed of electricity generation from sunlight; BC j This represents the upper limit of the battery capacity of the j-th photovoltaic-powered camera deployed on the highway.

[0027] Furthermore, the calculation formula for the feedback demand assessment value of each deployed billboard in S3 is as follows:

[0028] ;

[0029] Among them, FR j Let W represent the feedback demand assessment value of the j-th deployed billboard; let W denote the maximum wind speed of the j-th deployed billboard based on the current time's weather forecast information. j ; Obtain the maximum wind speed value corresponding to two consecutive images of the same billboard captured by the deployed photovoltaic camera; if there is no violation in the previous billboard image but a violation in the subsequent billboard image, bind the violation event corresponding to the subsequent billboard image with the maximum wind speed value corresponding to the two capture periods; SP j This indicates that within the set of violation events corresponding to the images captured by the j-th billboard deployed in the historical data, the bound wind speed is less than or equal to W. j A collection of billboard images related to various violations; SE j This indicates that within the set of violation events corresponding to the images captured by the j-th billboard deployed in the historical data, the bound wind speed is less than or equal to W. j The collection of billboard images associated with each semantic violation event; Len{} represents a function to count the number of elements in the collection; Num j This indicates that in the historical data, the maximum wind speed corresponding to the time interval between the captured image of the j-th billboard and the time of the previous image capture is less than or equal to W. j The total number of shots taken.

[0030] This invention analyzes the feedback demand assessment value of each deployed billboard by analyzing the percentage of images involving violations. During the analysis, the maximum wind speed during two consecutive shots of the same billboard by the photovoltaic camera is used as a filtering condition. This is because wind speeds on highways may carry surrounding objects or damage the billboard itself, leading to a risk of damage (involving image violations). Furthermore, considering that the violation analysis of billboard images involves analyzing not only image violations but also semantic violations for the same image, SP (Special Context) data needs to be obtained before using wind speed as a filtering condition. j and SE j The number of billboard images contained in the collection is then used to accurately calculate the feedback demand assessment value for each deployed billboard.

[0031] Furthermore, in the process of obtaining the shooting cycle adjustment coefficient corresponding to each photovoltaic-powered camera deployed on the highway in S4, the shooting cycle adjustment coefficient corresponding to the j-th photovoltaic-powered camera deployed on the highway is denoted as g. j ,

[0032] ;

[0033] Where, Sigmoid() represents the Sigmoid function; ξ represents the preset weight coefficient; EL j This represents the battery life risk assessment value corresponding to the j-th photovoltaic-powered camera deployed on the highway; FR j This represents the feedback demand assessment value of the j-th deployed billboard; the g j The range of values ​​for is (-1, 1);

[0034] The shooting cycle update result for each photovoltaic-powered camera deployed on the highway is equal to the product of the current shooting cycle of the corresponding photovoltaic-powered camera multiplied by 1 and the difference between the current shooting cycle and the shooting cycle adjustment coefficient of the corresponding photovoltaic-powered camera; the shooting cycle of the photovoltaic-powered camera is updated once every preset time interval.

[0035] A billboard content security monitoring system for highways includes an advertising image acquisition module, an advertising violation event analysis module, a battery life crisis and demand assessment and analysis module, and a shooting cycle adjustment and management module.

[0036] The advertising image acquisition module obtains the location information of billboards set up on the highway; the deployed photovoltaic-powered camera periodically takes pictures of the billboards set up on the highway, and the camera communicates with the cloud recognition platform through the 5G network to upload the location information of the corresponding camera and the billboard images taken each time;

[0037] The advertising violation event analysis module controls the cloud recognition platform to receive billboard images uploaded by the camera, extracts image features from the billboard images through image recognition technology, analyzes the content compliance of the billboard images, identifies the set of violation events corresponding to the billboard images, and binds the identified violation events with the corresponding billboard images and feeds them back to the display end.

[0038] The battery life crisis and demand assessment and analysis module obtains the shooting environment information and remaining power value of the photovoltaic power-powered cameras deployed on the highway at the current time; combines the current weather forecast information for each photovoltaic power-powered camera deployed on the highway to assess the battery life crisis coefficient corresponding to each photovoltaic power-powered camera deployed on the highway; and combines the set of violation events corresponding to the images captured by each deployed billboard in historical data to calculate the feedback demand assessment value of each deployed billboard.

[0039] The shooting cycle adjustment management module generates a shooting cycle adjustment coefficient for each photovoltaic-powered camera deployed on the highway based on the battery life crisis coefficient of each photovoltaic-powered camera deployed on the highway and the feedback demand assessment value of each billboard deployed on the highway, and updates the shooting cycle of each photovoltaic-powered camera deployed on the highway in real time.

[0040] Furthermore, the range crisis and demand assessment and analysis module includes a range crisis coefficient calculation unit and a feedback demand assessment and analysis unit.

[0041] The battery life crisis coefficient calculation unit obtains the shooting environment information and remaining power value of the photovoltaic power-powered cameras deployed on the highway at the current time; and combines the current weather forecast information for each photovoltaic power-powered camera deployed on the highway to evaluate the battery life crisis coefficient corresponding to each photovoltaic power-powered camera deployed on the highway.

[0042] The feedback demand assessment and analysis unit combines the set of violation events of the corresponding images of each deployed billboard in the historical data to calculate the feedback demand assessment value of each deployed billboard.

[0043] Furthermore, the shooting cycle adjustment management module includes a shooting cycle adjustment coefficient calculation unit and a shooting cycle dynamic control unit.

[0044] The shooting cycle adjustment coefficient calculation unit generates the shooting cycle adjustment coefficient for each photovoltaic-powered camera deployed on the highway based on the battery life crisis coefficient corresponding to each photovoltaic-powered camera deployed on the highway and the feedback demand assessment value of each billboard deployed.

[0045] The dynamic shooting cycle control unit updates the shooting cycle of each photovoltaic-powered camera deployed on the highway in real time according to the shooting cycle adjustment coefficient of each photovoltaic-powered camera deployed on the highway.

[0046] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: The present invention takes into account the impact of changes in the surrounding environment of the billboard on the photovoltaic power supply status of the camera, and solves the problem of camera power supply and battery life under continuous severe weather conditions; and realizes dynamic adjustment of the camera shooting cycle according to the remaining power and the future weather changes around the billboard, effectively extending the security monitoring time of the camera on the billboard content of the highway, and realizing effective supervision of the security of the billboard content of the highway. Attached Figure Description

[0047] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0048] Figure 1 This is a schematic diagram of the structure of the billboard content safety monitoring system for highways according to the present invention;

[0049] Figure 2 This is a flowchart illustrating the method for monitoring the safety of billboard content on highways according to the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0051] Please see Figures 1-2 The present invention provides a technical solution: such as Figure 1 As shown, this embodiment provides a billboard content security monitoring system for highways. The system includes an advertising image acquisition module, an advertising violation event analysis module, a battery life crisis and demand assessment and analysis module, and a shooting cycle adjustment and management module.

[0052] The advertising image acquisition module obtains the location information of billboards set up on the highway; the deployed photovoltaic-powered camera periodically takes pictures of the billboards set up on the highway, and the camera communicates with the cloud recognition platform through the 5G network to upload the location information of the corresponding camera and the billboard images taken each time;

[0053] The advertising violation event analysis module controls the cloud recognition platform to receive billboard images uploaded by the camera, extracts image features from the billboard images through image recognition technology, analyzes the content compliance of the billboard images, identifies the set of violation events corresponding to the billboard images, and binds the identified violation events with the corresponding billboard images and feeds them back to the display end.

[0054] The range crisis and demand assessment and analysis module includes a range crisis coefficient calculation unit and a feedback demand assessment and analysis unit.

[0055] The battery life crisis coefficient calculation unit obtains the shooting environment information and remaining power value of the photovoltaic power-powered cameras deployed on the highway at the current time; and combines the current weather forecast information for each photovoltaic power-powered camera deployed on the highway to evaluate the battery life crisis coefficient corresponding to each photovoltaic power-powered camera deployed on the highway.

[0056] The feedback demand assessment and analysis unit combines the set of violation events of the corresponding images of each deployed billboard in the historical data to calculate the feedback demand assessment value of each deployed billboard.

[0057] The shooting cycle adjustment management module includes a shooting cycle adjustment coefficient calculation unit and a shooting cycle dynamic control unit.

[0058] The shooting cycle adjustment coefficient calculation unit generates the shooting cycle adjustment coefficient for each photovoltaic-powered camera deployed on the highway based on the battery life crisis coefficient corresponding to each photovoltaic-powered camera deployed on the highway and the feedback demand assessment value of each billboard deployed.

[0059] The dynamic shooting cycle control unit updates the shooting cycle of each photovoltaic-powered camera deployed on the highway in real time according to the shooting cycle adjustment coefficient of each photovoltaic-powered camera deployed on the highway.

[0060] like Figure 2 As shown, this embodiment provides a method for security monitoring of billboard content on highways, the method comprising the following steps:

[0061] S1. Obtain the location information of billboards set up on the highway; periodically take pictures of billboards set up on the highway using deployed photovoltaic-powered cameras, wherein the cameras communicate with the cloud recognition platform through a 5G network and upload the location information of the corresponding cameras and the billboard images taken each time;

[0062] The shooting time period for each photovoltaic-powered camera deployed in S1 to capture images of billboards set up on the corresponding highway is obtained by querying the database; different cameras correspond to different shooting time periods.

[0063] During the communication process between the camera and the cloud recognition platform via the 5G network, the uploaded data also includes the time of each shot taken by the corresponding camera, and the billboard image captured each time is bound with the corresponding shooting time and the corresponding camera location information.

[0064] S2. The cloud-based recognition platform receives billboard images uploaded by cameras, extracts image features from the billboard images using image recognition technology, performs content compliance analysis on the billboard images, identifies the set of violations corresponding to the billboard images, and binds the identified violations with the corresponding billboard images, feeding them back to the display end.

[0065] S2 includes:

[0066] S21. The cloud recognition platform receives the billboard images uploaded by the camera and assigns a number to each uploaded billboard image; the number corresponding to the i-th uploaded billboard image is denoted as Bi.

[0067] S22. Extract text information from the billboard image corresponding to Bi using OCR technology, summarize the keywords in the extracted text information, and obtain the image semantic feature set of the billboard image corresponding to Bi.

[0068] S23. Obtain the average gray value of each pixel in the billboard image corresponding to Bi as the gray value of the corresponding billboard image, denoted as HBi; obtain the set of billboard images uploaded by the camera to which Bi belongs within the most recent preset time period, denoted as the reference set of Bi; obtain the billboard image in the reference set of Bi whose absolute value of the difference between the corresponding gray value and HBi is less than the preset gray value threshold and is closest to the shooting time of Bi, denoted as the comparison image of Bi.

[0069] S24. Obtain the difference in grayscale values ​​between the corresponding pixels in the same position between the comparison image corresponding to Bi and the billboard image corresponding to Bi. Mark the positions of pixels whose grayscale value difference is greater than a preset value in the billboard image corresponding to Bi. Use the pixel marking results in the billboard image corresponding to Bi as the image feature set of the billboard image corresponding to Bi.

[0070] S25. Compare each keyword in the semantic feature set of the billboard image corresponding to Bi with the preset keyword library corresponding to the semantic violation. The set of keywords belonging to the preset keyword library corresponding to the semantic violation in the comparison results is taken as the semantic violation judgment set. When the semantic violation judgment set is empty, it is determined that there is no semantic violation event in the billboard image corresponding to Bi; otherwise, it is determined that there is a semantic violation event in the billboard image corresponding to Bi.

[0071] S26. Divide the pixels in the image feature set of the billboard image corresponding to Bi into groups with a pixel distance less than a preset pixel distance, and remove groups with fewer than a preset number of pixels in each group. Record the image feature set of the billboard image corresponding to Bi after performing the removal of groups as the image violation judgment set. When the image violation judgment set is empty, it is determined that there is no image violation event in the billboard image corresponding to Bi; otherwise, it is determined that there is an image violation event in the billboard image corresponding to Bi.

[0072] S27. The sum of the judgment results of the visual violation event and the judgment results of the semantic violation event of the billboard image corresponding to Bi is denoted as the violation event set of the billboard image corresponding to Bi.

[0073] S3. Obtain the current shooting environment information and remaining power value of the photovoltaic-powered cameras deployed on the highway; combine the current weather forecast information for each photovoltaic-powered camera deployed on the highway to assess the battery life crisis coefficient corresponding to each photovoltaic-powered camera deployed on the highway; combine the set of violation events corresponding to the images captured by each deployed billboard in historical data to calculate the feedback demand assessment value of each deployed billboard.

[0074] The shooting environment information in S3 includes the effective visibility of the environment at the time of shooting; the weather forecast information includes the effective visibility prediction value, solar irradiance prediction value and wind speed corresponding to each time point in the subsequent preset time period based on the current time.

[0075] The calculation formula involved in S3 for evaluating the battery life crisis factor of each photovoltaic-powered camera deployed on the highway is as follows:

[0076] ;

[0077] Among them, EL j This represents the battery life risk assessment value corresponding to the j-th photovoltaic-powered camera deployed on the highway; TG j TG represents the length of the shooting time interval corresponding to the current time weather forecast information for the j-th photovoltaic-powered camera deployed on the highway; j At any point in the corresponding time interval, the effective visibility prediction value of the j-th photovoltaic-powered camera deployed on the highway is greater than the first visibility threshold based on the meteorological forecast information of the current time.

[0078] In this embodiment, the time interval in which the predicted effective visibility value of the j-th photovoltaic power-powered camera deployed on the highway is less than or equal to the first visibility threshold in the current weather forecast information is taken as the shooting sleep time interval of the corresponding photovoltaic power-powered camera; during the shooting sleep time interval, the photovoltaic power-powered camera provides operation;

[0079] BP j H represents the remaining power value of the j-th photovoltaic-powered camera deployed on the highway at the current time; j TP represents the power consumption of the j-th photovoltaic-powered camera deployed on the highway during each shot at the current time; j This represents the shooting cycle of the j-th photovoltaic-powered camera deployed on the highway at the current time; TB j G represents the length of the time interval in which the predicted solar irradiance value of the j-th photovoltaic-powered camera deployed on the highway is greater than the preset solar irradiance threshold based on the current time weather forecast information; （j，t） TB jThe predicted solar irradiance at time point t within the corresponding time interval; β （j，t） TB j The effective visibility prediction value at time point t within the corresponding time interval is the solar irradiance loss coefficient bound in the database preset form; F{} represents the operation of querying the solar power generation speed corresponding to different solar irradiance in the database preset form; This indicates that the solar irradiance in the database's preset form is... The corresponding speed of electricity generation from sunlight; BC j This represents the upper limit of the battery capacity of the j-th photovoltaic-powered camera deployed on the highway.

[0080] The calculation formula for the feedback demand assessment value of each deployed billboard in S3 is as follows:

[0081] ;

[0082] Among them, FR j Let W represent the feedback demand assessment value of the j-th deployed billboard; let W denote the maximum wind speed of the j-th deployed billboard based on the current time's weather forecast information. j ; Obtain the maximum wind speed value corresponding to two consecutive images of the same billboard captured by the deployed photovoltaic camera; if there is no violation in the previous billboard image but a violation in the subsequent billboard image, bind the violation event corresponding to the subsequent billboard image with the maximum wind speed value corresponding to the two capture periods; SP j This indicates that within the set of violation events corresponding to the images captured by the j-th billboard deployed in the historical data, the bound wind speed is less than or equal to W. j A collection of billboard images related to various violations; SE j This indicates that within the set of violation events corresponding to the images captured by the j-th billboard deployed in the historical data, the bound wind speed is less than or equal to W. j The collection of billboard images associated with each semantic violation event; Len{} represents a function to count the number of elements in the collection; Num j This indicates that in the historical data, the maximum wind speed corresponding to the time interval between the captured image of the j-th billboard and the time of the previous image capture is less than or equal to W. j The total number of shots taken.

[0083] S4. Based on the battery life crisis coefficient of each photovoltaic-powered camera deployed on the highway and the feedback demand assessment value of each billboard deployed, generate the shooting cycle adjustment coefficient of each photovoltaic-powered camera deployed on the highway, and update the shooting cycle of each photovoltaic-powered camera deployed on the highway in real time.

[0084] In step S4, during the process of obtaining the shooting period adjustment coefficient for each photovoltaic-powered camera deployed on the highway, the shooting period adjustment coefficient for the j-th photovoltaic-powered camera deployed on the highway is denoted as g. j ,

[0085] ;

[0086] Where, Sigmoid() represents the Sigmoid function; ξ represents the preset weight coefficient; EL j This represents the battery life risk assessment value corresponding to the j-th photovoltaic-powered camera deployed on the highway; FR j This represents the feedback demand assessment value of the j-th deployed billboard; the g j The range of values ​​for is (-1, 1);

[0087] The shooting cycle update result for each photovoltaic-powered camera deployed on the highway is equal to the product of the current shooting cycle of the corresponding photovoltaic-powered camera multiplied by 1 and the difference between the current shooting cycle and the shooting cycle adjustment coefficient of the corresponding photovoltaic-powered camera; the shooting cycle of the photovoltaic-powered camera is updated once every preset time interval.

[0088] In this embodiment, the update and adjustment of the shooting cycle corresponding to the photovoltaic-powered camera is a continuous process. The update result of the shooting cycle corresponding to the photovoltaic-powered camera may become smaller, remain unchanged, or become larger. The multiple of the updated shooting cycle compared to the shooting cycle before the update is (0,2).

[0089] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0090] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring the safety of billboard content on highways, characterized in that, The method includes the following steps: S1. Obtain the location information of billboards set up on the highway; periodically take pictures of billboards set up on the highway using deployed photovoltaic-powered cameras, wherein the cameras communicate with the cloud recognition platform through a 5G network and upload the location information of the corresponding cameras and the billboard images taken each time; S2. The cloud-based recognition platform receives billboard images uploaded by cameras, extracts image features from the billboard images using image recognition technology, performs content compliance analysis on the billboard images, identifies the set of violations corresponding to the billboard images, and binds the identified violations with the corresponding billboard images, feeding them back to the display end. S3. Obtain the current shooting environment information and remaining power value of the photovoltaic-powered cameras deployed on the highway; combine the current weather forecast information for each photovoltaic-powered camera deployed on the highway to assess the battery life crisis coefficient corresponding to each photovoltaic-powered camera deployed on the highway; combine the set of violation events corresponding to the images captured by each deployed billboard in historical data to calculate the feedback demand assessment value of each deployed billboard. The shooting environment information in S3 includes the effective visibility of the environment at the time of shooting; the weather forecast information includes the effective visibility prediction value, solar irradiance prediction value and wind speed corresponding to each time point in the subsequent preset time period based on the current time. The calculation formula involved in S3 for evaluating the battery life crisis factor of each photovoltaic-powered camera deployed on the highway is as follows: ； Among them, EL j This represents the battery life risk assessment value corresponding to the j-th photovoltaic-powered camera deployed on the highway; TG j TG represents the length of the shooting time interval corresponding to the current time weather forecast information for the j-th photovoltaic-powered camera deployed on the highway; j At any given time point within the corresponding time interval, the predicted effective visibility value of the j-th photovoltaic-powered camera deployed on the highway, based on the current time's weather forecast information, is greater than the first visibility threshold; BP j H represents the remaining power value of the j-th photovoltaic-powered camera deployed on the highway at the current time; j TP represents the power consumption of the j-th photovoltaic-powered camera deployed on the highway during each shot at the current time; j This represents the shooting cycle of the j-th photovoltaic-powered camera deployed on the highway at the current time; TB j G represents the length of the time interval in which the predicted solar irradiance value of the j-th photovoltaic-powered camera deployed on the highway is greater than the preset solar irradiance threshold based on the current time weather forecast information; （j，t） TB j The predicted solar irradiance at time point t within the corresponding time interval; β （j，t） TB j The effective visibility prediction value at time point t within the corresponding time interval is the solar irradiance loss coefficient bound in the database preset form; F{} represents the operation of querying the solar power generation speed corresponding to different solar irradiance in the database preset form; This indicates that the solar irradiance in the database's preset form is... The corresponding speed of electricity generation from sunlight; BC j This represents the upper limit of the battery capacity of the j-th photovoltaic-powered camera deployed on the highway; S4. Based on the battery life crisis coefficient of each photovoltaic-powered camera deployed on the highway and the feedback demand assessment value of each billboard deployed, generate the shooting cycle adjustment coefficient of each photovoltaic-powered camera deployed on the highway, and update the shooting cycle of each photovoltaic-powered camera deployed on the highway in real time.

2. The method for monitoring the content of billboards on highways according to claim 1, characterized in that: The shooting time period for each photovoltaic-powered camera deployed in S1 to capture images of billboards set up on the corresponding highway is obtained by querying the database; different cameras correspond to different shooting time periods. During the communication process between the camera and the cloud recognition platform via the 5G network, the uploaded data also includes the time of each shot taken by the corresponding camera, and the billboard image captured each time is bound with the corresponding shooting time and the corresponding camera location information.

3. The method for monitoring the content of billboards on highways according to claim 1, characterized in that: S2 includes: S21. The cloud recognition platform receives the billboard images uploaded by the camera and assigns a number to each uploaded billboard image; the number corresponding to the i-th uploaded billboard image is denoted as Bi. S22. Extract text information from the billboard image corresponding to Bi using OCR technology, summarize the keywords in the extracted text information, and obtain the image semantic feature set of the billboard image corresponding to Bi. S23. Obtain the average gray value of each pixel in the billboard image corresponding to Bi as the gray value of the corresponding billboard image, denoted as HBi; obtain the set of billboard images uploaded by the camera to which Bi belongs within the most recent preset time period, denoted as the reference set of Bi; obtain the billboard image in the reference set of Bi whose absolute value of the difference between the corresponding gray value and HBi is less than the preset gray value threshold and is closest to the shooting time of Bi, denoted as the comparison image of Bi. S24. Obtain the difference in grayscale values ​​between the corresponding pixels in the same position between the comparison image corresponding to Bi and the billboard image corresponding to Bi. Mark the positions of pixels whose grayscale value difference is greater than a preset value in the billboard image corresponding to Bi. Use the pixel marking results in the billboard image corresponding to Bi as the image feature set of the billboard image corresponding to Bi. S25. Compare each keyword in the semantic feature set of the billboard image corresponding to Bi with the preset keyword library corresponding to the semantic violation. The set of keywords belonging to the preset keyword library corresponding to the semantic violation in the comparison results is taken as the semantic violation judgment set. When the semantic violation judgment set is empty, it is determined that there is no semantic violation event in the billboard image corresponding to Bi; otherwise, it is determined that there is a semantic violation event in the billboard image corresponding to Bi. S26. Divide the pixels in the image feature set of the billboard image corresponding to Bi into groups with a pixel distance less than a preset pixel distance, and remove groups with fewer than a preset number of pixels in each group. Record the image feature set of the billboard image corresponding to Bi after performing the removal of groups as the image violation judgment set. When the image violation judgment set is empty, it is determined that there is no image violation event in the billboard image corresponding to Bi; otherwise, it is determined that there is an image violation event in the billboard image corresponding to Bi. S27. The sum of the judgment results of the visual violation event and the judgment results of the semantic violation event of the billboard image corresponding to Bi is denoted as the violation event set of the billboard image corresponding to Bi.

4. The method for monitoring the content of billboards on highways according to claim 1, characterized in that: The calculation formula for the feedback demand assessment value of each deployed billboard in S3 is as follows: ； Among them, FR j Let W represent the feedback demand assessment value of the j-th deployed billboard; let W denote the maximum wind speed of the j-th deployed billboard based on the current time's weather forecast information. j ; Obtain the maximum wind speed value corresponding to two consecutive images of the same billboard captured by the deployed photovoltaic camera; if there is no violation in the previous billboard image but a violation in the subsequent billboard image, bind the violation event corresponding to the subsequent billboard image with the maximum wind speed value corresponding to the two capture periods; SP j This indicates that within the set of violation events corresponding to the images captured by the j-th billboard deployed in the historical data, the bound wind speed is less than or equal to W. j A collection of billboard images related to various violations; SE j This indicates that within the set of violation events corresponding to the images captured by the j-th billboard deployed in the historical data, the bound wind speed is less than or equal to W. j The collection of billboard images associated with each semantic violation event; Len{} represents a function to count the number of elements in the collection; Num j This indicates that in the historical data, the maximum wind speed corresponding to the time interval between the captured image of the j-th billboard and the time of the previous image capture is less than or equal to W. j The total number of shots taken.

5. The method for monitoring the content of billboards on highways according to claim 1, characterized in that: In step S4, during the process of obtaining the shooting period adjustment coefficient for each photovoltaic-powered camera deployed on the highway, the shooting period adjustment coefficient for the j-th photovoltaic-powered camera deployed on the highway is denoted as g. j , ； Where, Sigmoid() represents the Sigmoid function; ξ represents the preset weight coefficient; EL j This represents the battery life risk assessment value corresponding to the j-th photovoltaic-powered camera deployed on the highway; FR j This represents the feedback demand assessment value of the j-th deployed billboard; the g j The range of values ​​for is (-1, 1); The shooting cycle update result for each photovoltaic-powered camera deployed on the highway is equal to the product of the current shooting cycle of the corresponding photovoltaic-powered camera multiplied by 1 and the difference between the current shooting cycle and the shooting cycle adjustment coefficient of the corresponding photovoltaic-powered camera; the shooting cycle of the photovoltaic-powered camera is updated once every preset time interval.

6. A billboard content security monitoring system for highways, employing the billboard content security monitoring method for highways as described in any one of claims 1-5, characterized in that, The system includes an advertising image acquisition module, an advertising violation event analysis module, a battery life crisis and demand assessment and analysis module, and a shooting cycle adjustment and management module. The advertising image acquisition module obtains the location information of billboards set up on the highway; the deployed photovoltaic-powered camera periodically takes pictures of the billboards set up on the highway, and the camera communicates with the cloud recognition platform through the 5G network to upload the location information of the corresponding camera and the billboard images taken each time; The advertising violation event analysis module controls the cloud recognition platform to receive billboard images uploaded by the camera, extracts image features from the billboard images through image recognition technology, analyzes the content compliance of the billboard images, identifies the set of violation events corresponding to the billboard images, and binds the identified violation events with the corresponding billboard images and feeds them back to the display end. The battery life crisis and demand assessment and analysis module obtains the shooting environment information and remaining power value of the photovoltaic power-powered cameras deployed on the highway at the current time; combines the current weather forecast information for each photovoltaic power-powered camera deployed on the highway to assess the battery life crisis coefficient corresponding to each photovoltaic power-powered camera deployed on the highway; and combines the set of violation events corresponding to the images captured by each deployed billboard in historical data to calculate the feedback demand assessment value of each deployed billboard. The shooting cycle adjustment management module generates a shooting cycle adjustment coefficient for each photovoltaic-powered camera deployed on the highway based on the battery life crisis coefficient of each photovoltaic-powered camera deployed on the highway and the feedback demand assessment value of each billboard deployed on the highway, and updates the shooting cycle of each photovoltaic-powered camera deployed on the highway in real time.

7. The billboard content security monitoring system for highways according to claim 6, characterized in that: The range crisis and demand assessment and analysis module includes a range crisis coefficient calculation unit and a feedback demand assessment and analysis unit. The battery life crisis coefficient calculation unit obtains the shooting environment information and remaining power value of the photovoltaic power-powered cameras deployed on the highway at the current time; and combines the current weather forecast information of each photovoltaic power-powered camera deployed on the highway to evaluate the battery life crisis coefficient corresponding to each photovoltaic power-powered camera deployed on the highway. The feedback demand assessment and analysis unit combines the set of violation events of the corresponding images of each deployed billboard in the historical data to calculate the feedback demand assessment value of each deployed billboard.

8. The billboard content security monitoring system for highways according to claim 6, characterized in that: The shooting cycle adjustment management module includes a shooting cycle adjustment coefficient calculation unit and a shooting cycle dynamic control unit. The shooting cycle adjustment coefficient calculation unit generates the shooting cycle adjustment coefficient for each photovoltaic-powered camera deployed on the highway based on the battery life crisis coefficient corresponding to each photovoltaic-powered camera deployed on the highway and the feedback demand assessment value of each billboard deployed. The dynamic shooting cycle control unit updates the shooting cycle of each photovoltaic-powered camera deployed on the highway in real time according to the shooting cycle adjustment coefficient of each photovoltaic-powered camera deployed on the highway.

Images