Road network monitoring method based on visual large model

Abstract

The invention discloses a road network monitoring method based on a visual large model, relates to the field of road network monitoring, and solves the problem that only congestion conditions and congestion duration can be checked in a traffic road network monitoring mode at the present stage, and the road network monitoring method specifically comprises the following steps: training historical road network data corresponding to a to-be-monitored road network through the visual large model; obtaining a local road network training model corresponding to the road network to be monitored; acquiring real-time video data corresponding to the road network to be monitored, and analyzing the real-time video data corresponding to the road network to be monitored to obtain a vehicle density mean value and a vehicle distance sensitive mean value of the road network to be monitored from the initial moment to the current moment; analyzing the same vehicle corresponding to the to-be-monitored road network at different moments to obtain a vehicle sensitive value of the to-be-monitored road network from the initial moment to the current moment; and analyzing the comprehensive condition of the to-be-monitored road network to obtain the road network sensitivity level of the to-be-monitored road network, so that accurate analysis and intelligent analysis of the traffic road network condition are realized.

Description

Technical Field

[0001] The invention relates to the technical field of traffic management, and in particular to a road network monitoring method based on a large visual model. Background Art

[0002] Traffic management plays an important role in the economic development of cities and regions. Reasonable planning and management of traffic can improve the efficiency of urban operations, promote the development of transportation and logistics industries, and thus promote economic development. Traffic congestion will bring inconvenience and waste of time to travel, and will also pollute the environment. Strengthening traffic management can alleviate traffic congestion through measures such as formulating traffic regulations and optimizing road layout. Traffic management can reduce transportation costs and improve transportation efficiency. Efficient monitoring of traffic networks can bring more convenience and benefits to people's production and life.

[0003] However, at present, when monitoring the traffic network, only the congestion situation and congestion duration of the traffic network can be checked through the vehicle GPS, but the sensitivity level of different roads in the traffic network cannot be checked, that is, the probability of accidents on different roads in the traffic network cannot be known; To this end, the present invention proposes a road network monitoring method based on a large visual model. Summary of the invention

[0004] The purpose of the present invention is to propose a road network monitoring method based on a large visual model to solve the problems raised in the above background technology.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions: A road network monitoring method based on a large visual model, the road network monitoring method is as follows: Step S1, obtaining historical road network data corresponding to the road network to be monitored; wherein the historical road network data is specifically road monitoring videos of the road network to be monitored in the previous year; Step S2, training the historical road network data corresponding to the road network to be monitored by using the large visual model to obtain a local road network training model corresponding to the road network to be monitored; Step S3, obtaining real-time video data corresponding to the road network to be monitored, analyzing the real-time video data corresponding to the road network to be monitored, and obtaining the average vehicle density and the average vehicle spacing sensitivity of the road network to be monitored from the initial time to the current time; Step S4, analyzing the road network to be monitored corresponding to the same vehicle at different times, and obtaining the vehicle sensitivity value of the road network to be monitored from the initial time to the current time; Step S5, analyzing the comprehensive situation of the road network to be monitored to obtain the road network sensitivity level of the road network to be monitored.

[0006] Furthermore, the step S3 includes the following sub-steps: Step S31, acquiring real-time video data of the road network to be monitored, and segmenting the real-time video data according to time to obtain images of the road network to be monitored corresponding to different time periods; The real-time video data is the surveillance video from the initial moment to the current moment, and each moment is numbered as n, where n=1, 2, ..., z, where z is a positive integer; Step S32, identifying the road network image to be monitored at the initial moment, and obtaining a vehicle distribution map corresponding to the road network to be monitored at the initial moment; Step S33, identifying the license plate number corresponding to the vehicle profile at the initial moment, recording the successfully identified vehicles as normal vehicles, recording the vehicles whose license plate numbers are not successfully identified as abnormal vehicles, and counting the number of normal vehicles ZCCn and the number of abnormal vehicles YCCn; Step S34, calculating the vehicle density value CLMn corresponding to the road network to be monitored at the initial time by a formula, the specific formula is as follows: CLMn=(ZCCn+YCCn×α) / e, where α is the correction coefficient for abnormal vehicles and e is a natural constant; Step S35, record the direction along the road network to be monitored as the X-axis direction, record the direction perpendicular to the road network to be monitored as the Y-axis direction, and establish a road network coordinate system.

[0007] Furthermore, the step S3 further includes the following sub-steps: Step S36, reading the inter-vehicle distance between two adjacent vehicles in the same lane, and comparing the inter-vehicle distance with the inter-vehicle distance threshold. If the inter-vehicle distance is greater than the inter-vehicle distance threshold, no operation is performed; if the inter-vehicle distance is less than or equal to the inter-vehicle distance threshold, the value of the inter-vehicle distance sensitivity coefficient at the corresponding moment is increased by one; Step S37, performing the operation of step S32 on the detected road network image at any time, obtaining the vehicle distribution map at the corresponding time, and then identifying the license plate number corresponding to the vehicle profile at different times, and calculating the vehicle density value at the corresponding time; Step S38, similarly reading the inter-vehicle distance between two adjacent vehicles in the same lane at any time, and obtaining the inter-vehicle distance sensitivity coefficient at the corresponding time; Step S39, adding and averaging multiple vehicle density values ​​to obtain the vehicle density mean CLMj from the initial moment to the current moment, and adding and averaging multiple vehicle spacing sensitivity coefficients to obtain the vehicle spacing sensitivity mean CJJj from the initial moment to the current moment.

[0008] Furthermore, the step S32 includes the following sub-steps: Step S321, obtaining the RGB pixel value of the road network image to be monitored corresponding to the initial moment; Step S322, obtaining an R value pixel map, a G value pixel map, and a B value pixel map corresponding to the road network image to be monitored according to the R value component, the G value component, and the B value component in the RGB pixel value respectively; Step S323, performing noise reduction filtering on the R value pixel map, the G value pixel map and the B value pixel map corresponding to the road network image to be monitored, respectively, to obtain the corresponding R value filter map, the G value filter map and the B value filter map; Step S324, combining the R value filter map, the G value filter map, and the B value filter map to obtain a filter noise reduction map corresponding to the initial moment; Step S325, importing the filtered denoising image into the local road network training model, automatically identifying and obtaining the vehicle distribution map and vehicle profile corresponding to the road network to be monitored at the initial moment.

[0009] Furthermore, the step S4 includes the following sub-steps: Step S41, obtaining a vehicle distribution map corresponding to different moments, and then obtaining a license plate number corresponding to a vehicle profile at each moment; Step S42, binding vehicles corresponding to the same license plate number, and recording the vehicle number as m, where m=1, 2, ..., p, where p is a positive integer; Step S43, for any vehicle, a vehicle travel coordinate diagram is established with the direction of the road network to be monitored as the X-axis and the direction perpendicular to the road network to be monitored as the Y-axis; the positions of the same vehicle at different times are plotted in the vehicle travel coordinate diagram; Step S44, for any vehicle, counting the total number of lane changes BDCm of the vehicle between the initial time and the current time; Step S45, the total number of lane changes of all vehicles is obtained in the same manner, and the total number of lane changes of any vehicle is compared with the lane change number threshold.

[0010] Furthermore, the process of determining whether a lane change occurs is as follows: If the same vehicle is in different lanes at two adjacent moments, it is deemed that the corresponding vehicle has changed lanes once.

[0011] Furthermore, the step S4 further includes the following sub-steps: Step S46: if the total number of lane changes of the vehicle is less than the lane change number threshold, no operation is performed; If the total number of lane changes of the vehicle is greater than or equal to the lane change number threshold, the corresponding vehicle is recorded as a lane change sensitive vehicle; The number of lane change sensitive vehicles is counted and recorded as lane change sensitive vehicle number BMC; Step S47, obtaining the number of abnormal vehicles corresponding to the initial time to the current time; Step S48, adding and averaging the number of abnormal vehicles from the initial moment to the current moment to obtain the average number of abnormal vehicles YCCp; Step S49, calculate the vehicle sensitivity value CMG corresponding to the road network to be monitored by a formula, and the calculation formula is as follows: CMG=YCCp×A1+BMC×A2; where A1 and A2 are proportional coefficients with fixed values, the values ​​of A1 and A2 are both greater than zero and A1+A2=1.

[0012] Furthermore, the step S5 includes the following sub-steps: Step S51, obtaining vehicle density values ​​corresponding to multiple time periods of the road network to be monitored, reading the maximum value among the vehicle density values ​​and recording it as the vehicle density peak value CLMf; Step S52, then obtaining the vehicle density mean CLMj, the vehicle spacing sensitivity mean CJJj and the vehicle sensitivity value CMG corresponding to the road network to be monitored; Step S53, calculating the road network sensitivity level value LMD corresponding to the road network to be monitored by a formula, the calculation formula is as follows: LMD=B1×(CLMf-CLMj)+B2×CJJj+B3×CMG; where B1, B2 and B3 are proportional coefficients with fixed values, and the values ​​of B1, B2 and B3 are all greater than zero; Step S54: if the road network sensitivity level value is greater than the first road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as level 3 sensitivity; If the road network sensitivity level value is less than or equal to the first road network sensitivity level threshold, and the road network sensitivity level value is greater than the second road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as secondary sensitivity; If the road network sensitivity level value is less than or equal to the second road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as level one sensitivity; Step S55, sending the road network sensitivity level corresponding to the road network to be monitored to the display terminal to inform the user of the corresponding situation of the road network to be monitored.

[0013] Further, the second road network sensitivity level threshold is less than the first road network sensitivity level threshold; The probability of abnormal events occurring in the road network to be monitored corresponding to the third-level sensitivity is greater than the probability of abnormal events occurring in the road network to be monitored corresponding to the second-level sensitivity; The probability of abnormal events occurring in the road network to be monitored corresponding to the second-level sensitivity is greater than the probability of abnormal events occurring in the road network to be monitored corresponding to the first-level sensitivity.

[0014] Furthermore, abnormal events include traffic jams, vehicle scratches, or the appearance of illegal vehicles.

[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: The present invention first obtains historical road network data corresponding to the road network to be monitored, trains the historical road network data corresponding to the road network to be monitored through a large visual model, thereby obtaining a local road network training model corresponding to the road network to be monitored, and then obtains real-time video data corresponding to the road network to be monitored, analyzes the real-time video data corresponding to the road network to be monitored, and obtains the average vehicle density and the average vehicle spacing sensitivity of the road network to be monitored from the initial moment to the current moment. At the same time, the road network to be monitored is analyzed corresponding to the same vehicle at different moments to obtain the vehicle sensitivity value of the road network to be monitored from the initial moment to the current moment. The comprehensive situation of the road network to be monitored is analyzed in combination with the average vehicle density, the average vehicle spacing sensitivity and the vehicle sensitivity value to obtain the road network sensitivity level of the road network to be monitored. The present invention realizes accurate and intelligent analysis of traffic road network conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0017] Figure 1 is a flow chart of the method of the present invention; Figure 2 A schematic diagram of filtering and noise reduction in the present invention; Figure 3 Schematic diagram of the road network coordinate system in the present invention; Figure 4 It is the vehicle driving coordinate diagram of the present invention.

[0018] Figure 5 It is a schematic diagram of the structure of the computer device in the present invention. DETAILED DESCRIPTION

[0019] The technical scheme of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] Embodiment 1: See also Figure 1-Figure 4 As shown, the technical solution provided by the present invention is: a road network monitoring method based on a visual large model, the method is specifically as follows: Step S1, obtaining historical road network data corresponding to the road network to be monitored; The historical road network data specifically refers to the road monitoring videos of the road network to be monitored in the previous year. In practice, with permission, the data can be obtained by accessing the cameras installed in the area where the road network to be monitored is located.

[0021] Step S2, training the historical road network data corresponding to the road network to be monitored by using the large visual model to obtain a local road network training model corresponding to the road network to be monitored; In this embodiment, the large visual model is a prior art, and the training process of the model is also common knowledge, so it is not described in detail in this embodiment.

[0022] Step S3, obtaining real-time video data corresponding to the road network to be monitored, analyzing the real-time video data corresponding to the road network to be monitored, and obtaining the average vehicle density and the average vehicle spacing sensitivity of the road network to be monitored from the initial time to the current time; In this embodiment, step S3 specifically includes the following sub-steps: Step S31, acquiring real-time video data of the road network to be monitored, and segmenting the real-time video data according to time to obtain images of the road network to be monitored corresponding to different time periods; The real-time video data specifically refers to the surveillance video from the initial moment to the current moment, and each moment is numbered as n, where n=1, 2, ..., z, where z is a positive integer; Step S32, identifying the road network image to be monitored at the initial moment, and obtaining a vehicle distribution map corresponding to the road network to be monitored at the initial moment; The step S32 includes the following sub-steps: Step S321, obtaining the RGB pixel value of the road network image to be monitored corresponding to the initial moment; Step S322, obtaining an R value pixel map, a G value pixel map, and a B value pixel map corresponding to the road network image to be monitored according to the R value component, the G value component, and the B value component in the RGB pixel value respectively; Step S323, performing noise reduction filtering on the R value pixel map, the G value pixel map and the B value pixel map corresponding to the road network image to be monitored, respectively, to obtain the corresponding R value filter map, the G value filter map and the B value filter map; The specific operation of Gaussian filtering is as follows: convolve the R value pixel map, G value pixel map and B value pixel map corresponding to the road network image to be monitored with the noise reduction kernel, and regard the noise reduction kernel as a sliding window, sliding in the R value pixel map, G value pixel map and B value pixel map respectively; perform point-to-point multiplication on the part where the pixel value overlaps with the sliding window and add the results to obtain the filter value at the corresponding position; like Figure 2 As shown in the figure, a certain R value pixel image is convolved with the noise reduction kernel, and the first pixel point of the corresponding R value filter image is converted to: 1×1+5×1+11×5+13×13=230 after noise reduction; the same applies to other pixel points; Step S324, combining the R value filter map, the G value filter map, and the B value filter map to obtain a filter noise reduction map corresponding to the initial moment; Step S325, importing the filtered denoising image into the local road network training model, automatically identifying and obtaining the vehicle distribution map and vehicle profile corresponding to the road network to be monitored at the initial moment; Step S33, identifying the license plate number corresponding to the vehicle profile at the initial moment, recording the successfully identified vehicles as normal vehicles, recording the vehicles whose license plate numbers are not successfully identified as abnormal vehicles, and counting the number of normal vehicles ZCCn and the number of abnormal vehicles YCCn; In specific operations, the text is automatically recognized using a camera equipped with an OCR algorithm. The basic principle of the OCR algorithm is to convert the text in the image into text data that can be edited and processed by a computer through a series of complex image processing and pattern recognition algorithms. Step S34, calculating the vehicle density value CLMn corresponding to the road network to be monitored at the initial time by a formula, the specific formula is as follows: CLMn=(ZCCn+YCCn×α) / e, where α is the correction coefficient for abnormal vehicles and e is a natural constant; Step S35, as Figure 3 As shown, the direction along the road network to be monitored is recorded as the X-axis direction, and the direction perpendicular to the road network to be monitored is recorded as the Y-axis direction, and a road network coordinate system is established; Step S36, reading the inter-vehicle distance between two adjacent vehicles in the same lane, and comparing the inter-vehicle distance with the inter-vehicle distance threshold. If the inter-vehicle distance is greater than the inter-vehicle distance threshold, no operation is performed; if the inter-vehicle distance is less than or equal to the inter-vehicle distance threshold, the value of the inter-vehicle distance sensitivity coefficient at the corresponding moment is increased by one; Step S37, performing the operation of step S32 on the detected road network image at any time, obtaining the vehicle distribution map at the corresponding time, and then identifying the license plate number corresponding to the vehicle profile at different times, and calculating the vehicle density value at the corresponding time; Step S38, similarly reading the inter-vehicle distance between two adjacent vehicles in the same lane at any time, and obtaining the inter-vehicle distance sensitivity coefficient at the corresponding time; Step S39, adding and averaging the multiple vehicle density values ​​to obtain the vehicle density mean CLMj from the initial moment to the current moment; The multiple vehicle spacing sensitivity coefficients are added and averaged to obtain the vehicle spacing sensitivity mean CJJj from the initial moment to the current moment.

[0023] Step S4, analyzing the road network to be monitored corresponding to the same vehicle at different times, and obtaining the vehicle sensitivity value of the road network to be monitored from the initial time to the current time; In this embodiment, step S4 specifically includes the following sub-steps: Step S41, obtaining a vehicle distribution map corresponding to different moments, and then obtaining a license plate number corresponding to a vehicle profile at each moment; Step S42, binding vehicles corresponding to the same license plate number, and recording the vehicle number as m, where m=1, 2, ..., p, where p is a positive integer; Step S43, for any vehicle, a vehicle travel coordinate diagram is established with the direction of the road network to be monitored as the X-axis and the direction perpendicular to the road network to be monitored as the Y-axis; the positions of the same vehicle at different times are plotted in the vehicle travel coordinate diagram, such as Figure 4 As shown; Step S44, for any vehicle, counting the total number of lane changes BDCm of the vehicle between the initial time and the current time; The specific process of determining whether a lane change has occurred is as follows: if the same vehicle is in different lanes at two adjacent moments, it is determined that the corresponding vehicle has changed lanes once; Step S45, the total number of lane changes of all vehicles is obtained by analogy, and the total number of lane changes of any vehicle is compared with the lane change number threshold; Step S46: if the total number of lane changes of the vehicle is less than the lane change number threshold, no operation is performed; If the total number of lane changes of the vehicle is greater than or equal to the lane change number threshold, the corresponding vehicle is recorded as a lane change sensitive vehicle; the number of lane change sensitive vehicles is counted and recorded as the lane change sensitive vehicle count BMC; Step S47, obtaining the number of abnormal vehicles corresponding to the initial time to the current time; Step S48, adding and averaging the number of abnormal vehicles from the initial moment to the current moment to obtain the average number of abnormal vehicles YCCp; Step S49, calculate the vehicle sensitivity value CMG corresponding to the road network to be monitored by a formula, and the calculation formula is as follows: CMG=YCCp×A1+BMC×A2; where A1 and A2 are proportional coefficients with fixed values, the values ​​of A1 and A2 are both greater than zero and A1+A2=1.

[0024] Step S5, analyzing the comprehensive situation of the road network to be monitored to obtain the road network sensitivity level of the road network to be monitored; In specific implementation, step S5 includes the following sub-steps: Step S51, obtaining vehicle density values ​​corresponding to multiple time periods of the road network to be monitored, reading the maximum value among the vehicle density values ​​and recording it as the vehicle density peak value CLMf; Step S52, obtaining the vehicle density mean CLMj and the vehicle spacing sensitivity mean CJJj corresponding to the road network to be monitored; obtaining the vehicle sensitivity value CMG corresponding to the road network to be monitored; Step S53, calculating the road network sensitivity level value LMD corresponding to the road network to be monitored by a formula, the calculation formula is as follows: LMD=B1×(CLMf-CLMj)+B2×CJJj+B3×CMG; where B1, B2 and B3 are proportional coefficients with fixed values, and the values ​​of B1, B2 and B3 are all greater than zero; Step S54: if the road network sensitivity level value is greater than the first road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as level 3 sensitivity; If the road network sensitivity level value is less than or equal to the first road network sensitivity level threshold, and the road network sensitivity level value is greater than the second road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as secondary sensitivity; If the road network sensitivity level value is less than or equal to the second road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as level one sensitivity; Among them, the second road network sensitivity level threshold is less than the first road network sensitivity level threshold; the probability of abnormal events occurring in the road network to be monitored corresponding to the third-level sensitivity is greater than the probability of abnormal events occurring in the road network to be monitored corresponding to the second-level sensitivity; the probability of abnormal events occurring in the road network to be monitored corresponding to the second-level sensitivity is greater than the probability of abnormal events occurring in the road network to be monitored corresponding to the first-level sensitivity; abnormal events specifically include traffic congestion, vehicle scratches, and the appearance of illegal vehicles; Step S55, sending the road network sensitivity level corresponding to the road network to be monitored to the display terminal to inform the user of the corresponding situation of the road network to be monitored; In this application, if corresponding calculation formulas appear, the above calculation formulas are all dimensionless and take their numerical calculations. The weight coefficients, proportional coefficients and other coefficients in the formulas are set to a result value obtained by quantifying each parameter. The size of the weight coefficient and the proportional coefficient can be determined as long as it does not affect the proportional relationship between the parameter and the result value.

[0025] Embodiment 2: Figure 5 The following is a schematic diagram of the structure of a computer device, such as Figure 5As shown, the computer device may include: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. The processor may call the logic instructions in the memory to execute a road network monitoring method based on a visual large model, the method comprising: obtaining historical road network data corresponding to the road network to be monitored; training the historical road network data corresponding to the road network to be monitored through the visual large model to obtain a local road network training model corresponding to the road network to be monitored; obtaining real-time video data corresponding to the road network to be monitored, analyzing the real-time video data corresponding to the road network to be monitored, and obtaining the mean vehicle density and the mean vehicle spacing sensitivity of the road network to be monitored from the initial moment to the current moment; analyzing the same vehicle corresponding to the road network to be monitored at different moments to obtain the vehicle sensitivity value of the road network to be monitored from the initial moment to the current moment; analyzing the comprehensive situation of the road network to be monitored to obtain the road network sensitivity level of the road network to be monitored.

[0026] In addition, the logic instructions in the above-mentioned memory can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0027] On the other hand, the present application also provides a computer program product, which includes a computer program stored on a computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute a road network monitoring method based on a visual big model provided by the above methods, the method including: obtaining historical road network data corresponding to the road network to be monitored; training the historical road network data corresponding to the road network to be monitored through the visual big model to obtain a local road network training model corresponding to the road network to be monitored; obtaining real-time video data corresponding to the road network to be monitored, analyzing the real-time video data corresponding to the road network to be monitored, and obtaining the average vehicle density and the average vehicle spacing sensitivity of the road network to be monitored from the initial moment to the current moment; analyzing the same vehicle corresponding to the road network to be monitored at different moments to obtain the vehicle sensitivity value of the road network to be monitored from the initial moment to the current moment; analyzing the comprehensive situation of the road network to be monitored to obtain the road network sensitivity level of the road network to be monitored.

[0028] On the other hand, the present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the above-mentioned road network monitoring method based on a visual big model, the method comprising: obtaining historical road network data corresponding to the road network to be monitored; training the historical road network data corresponding to the road network to be monitored through the visual big model to obtain a local road network training model corresponding to the road network to be monitored; obtaining real-time video data corresponding to the road network to be monitored, analyzing the real-time video data corresponding to the road network to be monitored, and obtaining the average vehicle density and the average vehicle spacing sensitivity of the road network to be monitored from the initial moment to the current moment; analyzing the same vehicle corresponding to the road network to be monitored at different moments to obtain the vehicle sensitivity value of the road network to be monitored from the initial moment to the current moment; analyzing the comprehensive situation of the road network to be monitored to obtain the road network sensitivity level of the road network to be monitored.

[0029] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0030] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0031] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A road network monitoring method based on a large visual model, characterized in that: The road network monitoring method is as follows: Step S1, obtaining historical road network data corresponding to the road network to be monitored; wherein the historical road network data is specifically road monitoring videos of the road network to be monitored in the previous year; Step S2, training the historical road network data corresponding to the road network to be monitored by using the large visual model to obtain a local road network training model corresponding to the road network to be monitored; Step S3, obtaining real-time video data corresponding to the road network to be monitored, analyzing the real-time video data corresponding to the road network to be monitored, and obtaining the average vehicle density and the average vehicle spacing sensitivity of the road network to be monitored from the initial time to the current time; Step S4, analyzing the road network to be monitored corresponding to the same vehicle at different times, and obtaining the vehicle sensitivity value of the road network to be monitored from the initial time to the current time; Step S5, analyzing the comprehensive situation of the road network to be monitored to obtain the road network sensitivity level of the road network to be monitored.

2. A road network monitoring method based on a visual large model according to claim 1, characterized in that: The step S3 includes the following sub-steps: Step S31, acquiring real-time video data of the road network to be monitored, and segmenting the real-time video data according to time to obtain images of the road network to be monitored corresponding to different time periods; The real-time video data is the surveillance video from the initial moment to the current moment, and each moment is numbered as n, where n=1, 2, ..., z, where z is a positive integer; Step S32, identifying the road network image to be monitored at the initial moment, and obtaining a vehicle distribution map corresponding to the road network to be monitored at the initial moment; Step S33, identifying the license plate number corresponding to the vehicle profile at the initial moment, recording the successfully identified vehicles as normal vehicles, recording the vehicles whose license plate numbers are not successfully identified as abnormal vehicles, and counting the number of normal vehicles ZCCn and the number of abnormal vehicles YCCn; Step S34, calculating the vehicle density value CLMn corresponding to the road network to be monitored at the initial time by a formula, the specific formula is as follows: CLMn=(ZCCn+YCCn×α) / e, where α is the correction coefficient for abnormal vehicles and e is a natural constant; Step S35, record the direction along the road network to be monitored as the X-axis direction, record the direction perpendicular to the road network to be monitored as the Y-axis direction, and establish a road network coordinate system.

3. A road network monitoring method based on a visual large model according to claim 2, characterized in that: The step S3 further comprises the following sub-steps: Step S36, reading the inter-vehicle distance between two adjacent vehicles in the same lane, and comparing the inter-vehicle distance with the inter-vehicle distance threshold. If the inter-vehicle distance is greater than the inter-vehicle distance threshold, no operation is performed; if the inter-vehicle distance is less than or equal to the inter-vehicle distance threshold, the value of the inter-vehicle distance sensitivity coefficient at the corresponding moment is increased by one; Step S37, performing the operation of step S32 on the detected road network image at any time, obtaining the vehicle distribution map at the corresponding time, and then identifying the license plate number corresponding to the vehicle profile at different times, and calculating the vehicle density value at the corresponding time; Step S38, similarly reading the inter-vehicle distance between two adjacent vehicles in the same lane at any time, and obtaining the inter-vehicle distance sensitivity coefficient at the corresponding time; Step S39, adding and averaging multiple vehicle density values ​​to obtain the vehicle density mean CLMj from the initial moment to the current moment, and adding and averaging multiple vehicle spacing sensitivity coefficients to obtain the vehicle spacing sensitivity mean CJJj from the initial moment to the current moment.

4. A road network monitoring method based on a visual large model according to claim 2, characterized in that: The step S32 includes the following sub-steps: Step S321, obtaining the RGB pixel value of the road network image to be monitored corresponding to the initial moment; Step S322, obtaining an R value pixel map, a G value pixel map, and a B value pixel map corresponding to the road network image to be monitored according to the R value component, the G value component, and the B value component in the RGB pixel value respectively; Step S323, performing noise reduction filtering on the R value pixel map, the G value pixel map and the B value pixel map corresponding to the road network image to be monitored, respectively, to obtain the corresponding R value filter map, the G value filter map and the B value filter map; Step S324, combining the R value filter map, the G value filter map, and the B value filter map to obtain a filter noise reduction map corresponding to the initial moment; Step S325, importing the filtered denoising image into the local road network training model, automatically identifying and obtaining the vehicle distribution map and vehicle profile corresponding to the road network to be monitored at the initial moment.

5. The road network monitoring method based on a visual large model according to claim 3 is characterized in that: The step S4 includes the following sub-steps: Step S41, obtaining a vehicle distribution map corresponding to different moments, and then obtaining a license plate number corresponding to a vehicle profile at each moment; Step S42, binding vehicles corresponding to the same license plate number, and recording the vehicle number as m, where m=1, 2, ..., p, where p is a positive integer; Step S43, for any vehicle, a vehicle travel coordinate diagram is established with the direction of the road network to be monitored as the X-axis and the direction perpendicular to the road network to be monitored as the Y-axis; the positions of the same vehicle at different times are plotted in the vehicle travel coordinate diagram; Step S44, for any vehicle, counting the total number of lane changes BDCm of the vehicle between the initial time and the current time; Step S45, the total number of lane changes of all vehicles is obtained in the same manner, and the total number of lane changes of any vehicle is compared with the lane change number threshold.

6. A road network monitoring method based on a visual large model according to claim 5, characterized in that: The specific process of determining lane change is as follows: If the same vehicle is in different lanes at two adjacent moments, it is deemed that the corresponding vehicle has changed lanes once.

7. A road network monitoring method based on a visual large model according to claim 5, characterized in that: The step S4 also includes the following sub-steps: Step S46: if the total number of lane changes of the vehicle is less than the lane change number threshold, no operation is performed; If the total number of lane changes of the vehicle is greater than or equal to the lane change number threshold, the corresponding vehicle is recorded as a lane change sensitive vehicle; The number of lane change sensitive vehicles is counted and recorded as lane change sensitive vehicle number BMC; Step S47, obtaining the number of abnormal vehicles corresponding to the initial time to the current time; Step S48, adding and averaging the number of abnormal vehicles from the initial moment to the current moment to obtain the average number of abnormal vehicles YCCp; Step S49, calculate the vehicle sensitivity value CMG corresponding to the road network to be monitored by a formula, and the calculation formula is as follows: CMG=YCCp×A1+BMC×A2; where A1 and A2 are proportional coefficients with fixed values, the values ​​of A1 and A2 are both greater than zero and A1+A2=1.

8. A road network monitoring method based on a visual large model according to claim 7, characterized in that: The step S5 comprises the following sub-steps: Step S51, obtaining vehicle density values ​​corresponding to multiple time periods of the road network to be monitored, reading the maximum value among the vehicle density values ​​and recording it as the vehicle density peak value CLMf; Step S52, then obtaining the vehicle density mean CLMj, the vehicle spacing sensitivity mean CJJj and the vehicle sensitivity value CMG corresponding to the road network to be monitored; Step S53, calculating the road network sensitivity level value LMD corresponding to the road network to be monitored by a formula, the calculation formula is as follows: LMD=B1×(CLMf-CLMj)+B2×CJJj+B3×CMG; where B1, B2 and B3 are proportional coefficients with fixed values, and the values ​​of B1, B2 and B3 are all greater than zero; Step S54: if the road network sensitivity level value is greater than the first road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as level 3 sensitivity; If the road network sensitivity level value is less than or equal to the first road network sensitivity level threshold, and the road network sensitivity level value is greater than the second road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as secondary sensitivity; If the road network sensitivity level value is less than or equal to the second road network sensitivity level threshold, the road network sensitivity level of the corresponding road network to be monitored is recorded as level one sensitivity; Step S55, sending the road network sensitivity level corresponding to the road network to be monitored to the display terminal to inform the user of the corresponding situation of the road network to be monitored.

9. A road network monitoring method based on a visual large model according to claim 8, characterized in that: The second road network sensitivity level threshold is less than the first road network sensitivity level threshold; The probability of abnormal events occurring in the road network to be monitored corresponding to the third-level sensitivity is greater than the probability of abnormal events occurring in the road network to be monitored corresponding to the second-level sensitivity, and the probability of abnormal events occurring in the road network to be monitored corresponding to the second-level sensitivity is greater than the probability of abnormal events occurring in the road network to be monitored corresponding to the first-level sensitivity.

10. A road network monitoring method based on a visual large model according to claim 9, characterized in that: Abnormal events include traffic jams, vehicle scratches or the appearance of illegal vehicles.

Images