Traffic facility collection system and method based on image recognition and GPS

Abstract

The application relates to the technical field of image processing, and discloses a traffic facility collection system based on image recognition and GPS, which comprises a collection module, a sign extraction module, a sign recognition module and a calibration module, wherein the field data comprises vehicle driving data and path shooting data; and the information recognition result comprises roadside sign information and corresponding vehicle GPS position information. The traffic facility direction on both sides of a road can be collected and acquired by using a vehicle to carry shooting and GPS positioning equipment, the pictures related to the traffic facilities are automatically extracted, the information recognition result related to the traffic facilities is automatically acquired from the pictures, then a path calibration map containing a driving track, road information and roadside traffic facility information is generated, and the road network informatization work of relevant staff is facilitated.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and specifically to a traffic facility data acquisition system and method based on image recognition and GPS. Background Technology

[0002] The adjustment of the national highway network plan has created an urgent need for the survey and optimization of highway signs. The adjustment requires traffic facilities and markings to adapt to the new plan, meeting the operational and safety needs of roads with different functions. The renaming and numbering of expressways must be coordinated with the adjustment of traffic facilities, and the formulation and adjustment of standards and specifications must be based on the survey, evaluation, and optimization of existing traffic facilities.

[0003] With the development of concepts and technologies such as the Internet of Things (IoT), the Internet of Vehicles (IoV), and sensor networks, the level of intelligent transportation has been greatly improved. However, this has also placed higher demands on the intelligent informatization level of infrastructure and the management level of personnel. Regarding transportation facilities, existing navigation equipment can provide some information such as speed limits, speed monitoring, and road signs, but there are still problems with incomplete information and outdated updates. In the daily maintenance and management of transportation facilities, a complete database has not been established for the attributes, location, status, and relationship with the surrounding road environment of road transportation facilities. Additions, reductions, or adjustments to transportation facilities are also not reflected in the database, and this is only one of the most basic aspects of intelligent informatization of transportation facilities. Establishing a nationwide (national highway network) database of transportation facilities through a comprehensive survey is the foundation for the future informatization of transportation facilities.

[0004] Traffic facility surveys are a crucial task during the road operation and management phase. Traffic facilities are dynamically changing during operation and management, making surveys and optimization essential daily tasks. Currently, most cities rely primarily on manual methods for traffic facility surveys. This approach suffers from low efficiency, poor accuracy, cumbersome data statistical analysis, and heavy workloads in both fieldwork and office work. Another drawback is that the surveyed traffic facility information only includes station numbers, lacking other location information suitable for networking. Therefore, it's difficult to integrate this information across the entire road or network, and also difficult to conduct comparative analysis with other roads of the same grade. Furthermore, centralized surveys cannot adequately cover newly added or modified traffic facilities, and are uneconomical. Summary of the Invention

[0005] The purpose of this invention is to provide a traffic facility data collection system and method based on image recognition and GPS, and to solve the following technical problems:

[0006] How to provide a traffic facility data collection system that can improve the efficiency and accuracy of road network information work?

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] A traffic facility data collection system based on image recognition and GPS includes:

[0009] The data acquisition module is used to collect field data along designated roads;

[0010] A marker extraction module, connected to the acquisition module, is used to obtain marker extraction images based on the field data;

[0011] A sign recognition module, connected to the acquisition module and the sign extraction module, is used to output corresponding information recognition results based on the extracted sign image;

[0012] A calibration module, connected to the sign recognition module, is used to generate a path calibration map based on the field data and the information recognition results;

[0013] The field data includes vehicle driving data and route shooting data; the information recognition results include roadside sign information and corresponding vehicle GPS location information.

[0014] The above technical solution allows for the use of vehicles equipped with cameras and GPS positioning devices to collect field data on the direction of traffic facilities along roads. It automatically extracts images related to traffic facilities and automatically obtains information recognition results related to traffic facilities from the images. Then, it generates a route mapping map containing driving trajectory, road information, and roadside traffic facility information, which facilitates the road network informatization work of relevant personnel.

[0015] As a further aspect of the present invention: the path-captured data consists of road condition images acquired at a preset frequency, participating in... Figure 3 and Figure 4 As shown;

[0016] The flag extraction module includes:

[0017] The detection module is used to detect whether traffic facilities and designated reference objects exist in the road condition image; if traffic facilities are detected, the module outputs and displays Box1 of the traffic facilities; if the designated reference object is detected at the same time, the module outputs and displays Box2 of the designated reference object; if the designated reference object corresponding to Box2 appears for the first time, it is recorded as Box2S; S is the corresponding reference recognition result.

[0018] The inspection module is used to check whether the Box2 classification of two adjacent road condition images is consistent and both are S; if they are consistent, the previous road condition image is removed; otherwise, both road condition images are retained, and the new reference recognition result is associated with the corresponding road condition image.

[0019] As a further aspect of the present invention: for all Box1 corresponding to the same traffic facility, the inspection module is further configured to use an edge cutting algorithm to cut the traffic facility in the traffic image to obtain a closed sign outline Box1*;

[0020]

[0021] Extract the image using Box1 corresponding to IoU≥μ as the marker;

[0022] Where IoU is the crossover-union ratio, S Box1 S is the area enclosed by Box1. Box1* The area enclosed by Box1*; μ is a preset value.

[0023] As a further aspect of the present invention: the vehicle driving data includes vehicle driving information and vehicle GPS location information; the vehicle driving information includes vehicle speed and steering angle;

[0024] The preset frequency is positively correlated with the vehicle speed and the magnitude of the steering angle.

[0025] As a further aspect of the present invention: such as Figure 5 As shown, the path calibration map includes vehicle driving trajectory and calibration lines. The calibration lines are corrected according to the vehicle driving trajectory, and 100-meter markers are set along the calibration lines according to the vehicle driving information.

[0026] As a further aspect of the present invention, it also includes a cloud server and a console;

[0027] The cloud server is connected to the calibration module and is used to store and update the path calibration map;

[0028] The console is connected to the cloud server to support online use by multiple users and services, enabling unified display of various types of traffic facilities. Clicking on a road point that has collected data allows users to view the corresponding road photo.

[0029] A method for collecting traffic facility data based on image recognition and GPS, comprising:

[0030] Collect field data along the designated roads;

[0031] Based on the field data, extract the logo image;

[0032] Extract the image based on the logo and output the corresponding information recognition result;

[0033] A path calibration map is generated based on the field data and the information recognition results;

[0034] The field data includes vehicle driving data and route shooting data; the information recognition results include roadside sign information and corresponding vehicle GPS location information.

[0035] The beneficial effects of this invention are as follows: This invention uses vehicle-mounted camera and GPS positioning equipment to collect field data on the direction of traffic facilities on both sides of the road, automatically extracts images related to traffic facilities, and automatically obtains information recognition results related to traffic facilities from the images. Then, it generates a route mapping map containing driving trajectory, road information, and roadside traffic facility information, which facilitates the road network informatization work of relevant personnel. Attached Figure Description

[0036] The invention will now be further described with reference to the accompanying drawings.

[0037] Figure 1 This is a schematic diagram of the module connections of the traffic facility data acquisition system in this invention;

[0038] Figure 2 This is a schematic diagram illustrating the operation of the marker extraction module and the marker recognition module in this invention;

[0039] Figure 3 This is an example diagram of marker extraction in this invention;

[0040] Figure 4 These are road condition images for some sections of the road used in this invention;

[0041] Figure 5 This is a route mapping map for some road sections in this invention. Detailed Implementation

[0042] 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.

[0043] Please see Figure 1 As shown, the present invention is a traffic facility data collection system based on image recognition and GPS, comprising:

[0044] The data acquisition module is used to collect field data along designated roads;

[0045] A marker extraction module, connected to the acquisition module, is used to obtain marker extraction images based on the field data;

[0046] A sign recognition module, connected to the acquisition module and the sign extraction module, is used to output corresponding information recognition results based on the extracted sign image;

[0047] A calibration module, connected to the sign recognition module, is used to generate a path calibration map based on the field data and the information recognition results;

[0048] The field data includes vehicle driving data and route shooting data; the information recognition results include roadside sign information and corresponding vehicle GPS location information.

[0049] Through the above technical solution, this invention uses vehicle-mounted shooting and GPS positioning equipment to collect field data on the direction of traffic facilities on both sides of the road. The GPS can adopt the Beidou GPS 10mm specification, and a 4K high-definition industrial camera is used, which can achieve 4K image shooting at speeds up to 120km / h. At the same time, the heat dissipation problem is solved. It automatically extracts images related to traffic facilities, such as traffic signs, and automatically obtains the information recognition results related to traffic facilities from the images. Then, it generates a path calibration map containing driving trajectory, road information, and roadside traffic facility information, which facilitates the road network informatization work of relevant personnel.

[0050] As a further aspect of the present invention: the path shooting data is road condition images acquired at a preset frequency;

[0051] The flag extraction module includes:

[0052] The detection module is used to detect whether traffic facilities and designated reference objects exist in the road condition image; if traffic facilities are detected, the module outputs and displays Box1 of the traffic facilities; if the designated reference object is detected at the same time, the module outputs and displays Box2 of the designated reference object; if the designated reference object corresponding to Box2 appears for the first time, it is recorded as Box2S; S is the corresponding reference recognition result.

[0053] The inspection module is used to check whether the Box2 classification of two adjacent road condition images is consistent and both are S; if they are consistent, the previous road condition image is removed; otherwise, both road condition images are retained, and the new reference recognition result is associated with the corresponding road condition image.

[0054] like Figure 2 As shown, the inspection module is designed based on an AI model, which can detect the classification and similarity of objects within Box 2. The YOLO-based AI model can identify traffic facilities in road condition images. This single-target detection and accurate extraction of road sign areas in sampled images achieves efficient deduplication and improves the accuracy of AI recognition and judgment, reducing the pressure of subsequent manual annotation and review.

[0055] As a further aspect of the present invention: for all Box1 corresponding to the same traffic facility, the inspection module is further configured to use an edge cutting algorithm to cut the traffic facility in the traffic image to obtain a closed sign outline Box1*;

[0056]

[0057] Extract the image using Box1 corresponding to IoU≥μ as the marker;

[0058] Where IoU is the crossover-union ratio, S Box1 S is the area enclosed by Box1. Box1* The area enclosed by Box1*; μ is a preset value. In this embodiment of the invention, the preset value can be selected as 0.8-0.9.

[0059] As a further aspect of the present invention: the vehicle driving data includes vehicle driving information and vehicle GPS location information; the vehicle driving information includes vehicle speed and steering angle;

[0060] The preset frequency is positively correlated with the vehicle speed and the steering angle. In this embodiment of the invention, the preset frequency for taking road condition photos can be set to 1 photo per second. If the vehicle speed and steering angle increase, it can be adjusted to 2 or 3 photos per second according to preset rules. The specific frequency is set according to the actual situation.

[0061] As a further aspect of the present invention: the path calibration map includes vehicle driving trajectory and calibration line, the calibration line is corrected according to the vehicle driving trajectory, and 100-meter stakes are set along the calibration line according to the vehicle driving information.

[0062] As a further aspect of the present invention, it also includes a cloud server and a console;

[0063] The cloud server is connected to the calibration module and is used to store and update the path calibration map;

[0064] The console is connected to the cloud server to support online use by multiple users and services, enabling unified display of various types of traffic facilities. Clicking on a road point that has collected data allows users to view the corresponding road photo.

[0065] A method for collecting traffic facility data based on image recognition and GPS, comprising:

[0066] Collect field data along the designated roads;

[0067] Based on the field data, extract the logo image;

[0068] Extract the image based on the logo and output the corresponding information recognition result;

[0069] A path calibration map is generated based on the field data and the information recognition results;

[0070] The field data includes vehicle driving data and route shooting data; the information recognition results include roadside sign information and corresponding vehicle GPS location information.

[0071] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A traffic facility data collection system based on image recognition and GPS, characterized in that, include: The data acquisition module is used to collect field data along designated roads; A marker extraction module, connected to the acquisition module, is used to obtain marker extraction images based on the field data; A sign recognition module, connected to the acquisition module and the sign extraction module, is used to output corresponding information recognition results based on the extracted sign image; A calibration module, connected to the sign recognition module, is used to generate a path calibration map based on the field data and the information recognition results; The field data includes vehicle driving data and route shooting data; the information recognition results include roadside sign information and corresponding vehicle GPS location information. The path shooting data consists of road condition images acquired at a preset frequency; The flag extraction module includes: The detection module is used to detect whether traffic facilities and designated reference objects exist in the road condition image; if traffic facilities are detected, the module outputs and displays Box1 of the traffic facilities; if the designated reference object is detected at the same time, the module outputs and displays Box2 of the designated reference object; if the designated reference object corresponding to Box2 appears for the first time, it is recorded as Box2S; S is the corresponding reference recognition result. The inspection module is used to check whether the Box2 classification of two adjacent road condition images is consistent and both are S; if they are consistent, the previous road condition image is removed; otherwise, the two road condition images are retained, and the new reference recognition result is associated with the corresponding road condition image. For all Box1 corresponding to the same traffic facility, the inspection module is further configured to use an edge cutting algorithm to cut the traffic facility in the traffic image to obtain a closed sign outline Box1*. ； Will The corresponding Box1 is used as the image for the logo extraction; in, For intersection, union, and comparison, Let this be the area enclosed by Box1. The area enclosed by Box1*; This is a preset value.

2. The traffic facility data collection system based on image recognition and GPS according to claim 1, characterized in that, The vehicle driving data includes vehicle driving information and vehicle GPS location information; the vehicle driving information includes vehicle speed and steering angle. The preset frequency is positively correlated with the vehicle speed and the magnitude of the steering angle.

3. The traffic facility data collection system based on image recognition and GPS according to claim 2, characterized in that, The path calibration map includes vehicle driving trajectory and calibration lines. The calibration lines are corrected according to the vehicle driving trajectory, and 100-meter markers are set along the calibration lines according to the vehicle driving information.

4. The traffic facility data collection system based on image recognition and GPS according to claim 1, characterized in that, It also includes cloud servers and a console; The cloud server is connected to the calibration module and is used to store and update the path calibration map; The console is connected to the cloud server to support online use by multiple users and services, enabling unified display of various types of traffic facilities. Clicking on a road point that has collected data allows users to view the corresponding road photo.

Images