Target detection optimization method and device, electronic equipment and storage medium

Abstract

The invention relates to the technical field of target detection, in particular to a target detection optimization method and device, electronic equipment and a storage medium, and the method comprises the following steps: S1, image collection: preparing an image collection module which is used for carrying out the image collection and analysis of an image; s2, environment acquisition: preparing an environment acquisition module, wherein the environment acquisition module is used for acquiring radar images, environment temperature and environment humidity; s3, comprehensive analysis: preparing a comprehensive analysis module which is used for receiving and processing the image information, thermal imaging and target cloud points acquired by the image acquisition module and the environment acquisition module, and monitoring the driving state of the vehicle in real time; and S4, driving scheme analysis: constructing a vehicle automatic driving model based on a convolutional layer algorithm by using a deep learning module according to historical driving records of the vehicle in different driving environments. The system is complete in function, can effectively carry out target detection, and assists in vehicle driving.

Description

Technical Field

[0001] The present invention relates to the field of target detection technology, and in particular to a target detection optimization method, device, electronic equipment and storage medium. Background Art

[0002] The object detection optimization method is mainly used in autonomous driving technology, which is a technology that deeply integrates the automotive industry with new-generation information such as artificial intelligence, high-performance computing, big data, and the Internet of Things. It has significant advantages in reducing traffic congestion, reducing traffic accidents, and improving fuel economy. In addition, the development of autonomous driving technology can help cities build a safe and efficient future travel structure, which is of great strategic significance to the cross-border integration development of the automotive industry.

[0003] In the prior art, when performing image acquisition, target detection optimization methods usually use visible light cameras for acquisition, which have many advantages such as high resolution, low cost, stable working performance, etc. Visible light cameras can collect rich semantic information, color information and texture information from vehicle targets, but these features are easily disturbed by bad weather, resulting in a significant reduction in the accuracy of image acquisition by visible light cameras in bad weather.

[0004] In summary, how to solve the problem that the target detection optimization method in the prior art is easily disturbed by bad weather, resulting in a significant reduction in the accuracy of image acquisition in bad weather has become a difficult problem that needs to be solved in the current field. Therefore, it is necessary to propose a target detection optimization method, device, electronic device and storage medium that can improve the accuracy of image acquisition in bad weather. Summary of the invention

[0005] To solve the above problems, the present invention provides a target detection optimization method, device, electronic device and storage medium. Through the design of image acquisition module and environment acquisition module, the vehicle's environment can be effectively identified, and then further analyzed through the comprehensive analysis module to improve the accuracy of target detection.

[0006] In order to achieve the above object, the technical solution of the present invention is as follows: a target detection optimization method, comprising the following steps:

[0007] S1, image acquisition: prepare the image acquisition module, which is used to acquire and analyze the image of the detection target; use the image acquisition module to acquire the image information around the vehicle, analyze the weather and road conditions based on the image information, and select pedestrians, vehicles, traffic facilities and obstacles in the image.

[0008] S2, environment acquisition: prepare the environment acquisition module, which is used to collect radar images, environmental temperature and environmental humidity to form thermal imaging and target cloud points; use the environment acquisition module to collect information about the vehicle's surrounding environment; the environment acquisition module collects the temperature of each detection target, and displays the temperature image and temperature profile of the detection target in the form of thermal imaging. According to the temperature image and temperature profile of the detection target, a comparison and analysis is performed to determine the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment.

[0009] At the same time, the environment acquisition module is used to transmit electromagnetic wave signals from the vehicle's surrounding environment, receive echo signals, analyze and process the echo signals, extract the distance, size, shape, direction and speed of the detection target, judge the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment, and display them in the form of target cloud points;

[0010] The environment acquisition module will collect the environmental humidity and assist the image acquisition module in determining the current weather conditions.

[0011] S3, comprehensive analysis: prepare a comprehensive analysis module, which is used to receive and process the image information, thermal imaging and target cloud points collected by the image acquisition module and the environment acquisition module, and monitor the vehicle's driving status in real time; use the comprehensive analysis module to monitor the vehicle's driving status in real time, and combine the image information and environmental information to comprehensively analyze the pedestrians, vehicles, traffic facilities, road conditions and obstacles around the vehicle to derive the current driving environment of the vehicle, and record the vehicle's driving environment and driving status each time to form a historical driving record.

[0012] S4, driving plan analysis: prepare a deep learning module, which is used to perform deep learning based on historical driving records and optimize driving plans; use the deep learning module to build a vehicle automatic driving model based on the convolutional layer algorithm according to the vehicle's historical driving records in different driving environments; simulate different driving plans based on the automatic driving model, combined with the current vehicle driving status and driving environment, and analyze the driving plan with a low accident rate and compliance with road traffic regulations based on the driving conditions simulated by different driving plans, the vehicle's driving status, the vehicle's surrounding environment and road traffic regulations.

[0013] Furthermore, in S1, the image acquisition module identifies and distinguishes pedestrians, vehicles and traffic facilities based on their overall shape and contour features.

[0014] Furthermore, in S1, the road surface conditions include road surface type, road surface material, road surface markings and road surface unevenness.

[0015] Furthermore, in S1, the road surface types include straight sections, turning sections, uphill sections, and downhill sections.

[0016] Furthermore, in S1, the driving state includes the vehicle's driving speed, turning angle, and light usage.

[0017] Furthermore, in S1, obstacles include objects appearing on the road surface other than pedestrians, vehicles, traffic facilities and road surfaces.

[0018] Furthermore, in S2, the ambient temperature includes the ground surface temperature and the vehicle interior temperature.

[0019] The technical principles of the above scheme are as follows:

[0020] First, a large amount of image recognition training is performed on the image acquisition module, and the trained image acquisition module is used to collect image information around the vehicle. During the image acquisition process, the image acquisition module will select and display pedestrians, vehicles, traffic facilities and obstacles in the image with different color frames, thereby reminding the driver of the pedestrians, vehicles, traffic facilities and obstacles that need to be paid attention to around his vehicle.

[0021] While collecting images, the environment collection module is used to collect the temperature conditions of each detection target. Since the temperatures of pedestrians, vehicles, traffic facilities, road surfaces and obstacles are all different, the temperature of vehicle engines is much higher than the body temperature of pedestrians, while the temperatures of road surfaces and obstacles are generally lower in normal weather. Therefore, the environment collection module uses thermal imaging to display the temperature image and temperature profile of the detection target. The lower the temperature, the bluer the image color, and the higher the temperature, the redder the image color. As the temperature rises, the image will change from dark blue to dark red. The environment collection module then compares and analyzes the temperature image and temperature profile of the detection target with the temperature images and temperature profiles of pedestrians, vehicles, traffic facilities, road surfaces and obstacles under normal circumstances, thereby effectively judging the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment.

[0022] Since the distance, speed, shape and size of different objects are different, the reception and reflection of waves will also be different. Since the propagation speed of electromagnetic waves is known, its speed is about the speed of light in a vacuum. The radar can calculate the distance between the target and the radar by measuring the time delay between the transmitted signal and the received echo. Therefore, the environment acquisition module will also transmit electromagnetic wave signals from the vehicle's surrounding environment, receive echo signals, analyze and process the echo signals, extract the distance, size, shape, direction and speed of the detection target, and determine the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment; the environment acquisition module will determine the azimuth and elevation of the detection target relative to the radar through the joint scanning of multiple antennas, thereby determining the position of the detection target in three-dimensional space, and display the detection target in the form of target cloud points, so that the driver can understand the vehicle's surrounding environment more intuitively.

[0023] The environment acquisition module will collect the ambient humidity and transmit the ambient humidity data to the image acquisition module to assist the image acquisition module in determining the current weather conditions.

[0024] After the image acquisition and environment acquisition are completed, the comprehensive analysis module will extract the image acquisition and environment acquisition, give priority to judging the current weather conditions, and select the main judgment basis; if the weather conditions are hot, it is easy to cause the temperature difference between pedestrians, roads and obstacles to decrease during environment acquisition, making the temperature image in the thermal imaging similar in color and the temperature profile clarity reduced. At this time, the comprehensive analysis module will use the image information collected by the image acquisition module and the target cloud points collected by the radar as the main judgment basis to judge the pedestrians, vehicles, traffic facilities, road conditions and obstacles around the vehicle, and then use thermal imaging for auxiliary judgment; if it is raining, the image acquisition is prone to blurring, and the radar will also be affected. At this time, the vehicle engine temperature is still much higher than the pedestrian's body temperature, while the traffic facilities, road surface and obstacle temperatures will be lower than the pedestrian's body temperature. The comprehensive analysis module will use the temperature image and temperature profile collected by thermal imaging as the main judgment basis, and then use radar and image information for auxiliary judgment; in this way, the vehicle's surrounding environment can be detected under different weather conditions, achieving a more comprehensive detection effect.

[0025] The deep learning module will build a vehicle autonomous driving model based on the convolutional layer algorithm according to the vehicle's historical driving records in different driving environments; the autonomous driving model will simulate different driving schemes based on the current vehicle driving status and driving environment, and analyze the driving conditions simulated by different driving schemes based on the vehicle's driving status, the vehicle's surrounding environment and road traffic regulations to obtain a driving plan with a low accident rate and compliance with road traffic regulations.

[0026] The above scheme has the following beneficial effects:

[0027] 1. The present invention uses the image acquisition module to recognize detailed images of the vehicle's surrounding environment, especially the real-time selection and display of pedestrians, vehicles, traffic facilities and obstacles, which significantly improves the driver's perception of the surrounding environment and reduces the risk of accidents caused by blind spots or negligence. At the same time, the environment acquisition module uses thermal imaging technology to effectively identify potential dangerous factors, such as puddles and obstacles on the road, even in complex or severe weather conditions, further improving driving safety.

[0028] 2. The design of the comprehensive analysis module of the present invention can automatically adjust the judgment basis according to different weather conditions, ensuring that the vehicle's surrounding environment can be effectively detected in various adverse weather conditions. For example, in hot weather, when thermal imaging technology is limited by the reduction of temperature differences, the comprehensive analysis module will rely more on image information and radar data for judgment; on rainy days, when image and radar data are interfered with, the comprehensive analysis module will rely more on thermal imaging data. This flexible switching mechanism greatly enhances the environmental adaptability of the system.

[0029] 3. The design of the deep learning module of the present invention builds an automatic driving model by analyzing the historical driving records of the vehicle in different driving environments. The deep learning module can simulate different driving plans, evaluate their accident risks and legality, and finally select the best driving plan. This decision-making method based on big data and machine learning enables vehicles to respond to complex traffic environments more intelligently and improve driving safety and efficiency.

[0030] Furthermore, a target detection device includes a mounting bar, and a temperature sensor, a humidity sensor, an image collector and a radar are fixedly connected to the side wall of the mounting bar.

[0031] Beneficial effects: The design of the mounting bar can assemble the temperature sensor, humidity sensor, image collector and radar into a whole, so as to better install the target detection device on the vehicle, thereby providing better target detection service for the vehicle.

[0032] Furthermore, an electronic device for target detection includes a controller, which is used to receive information collected by a temperature sensor, a humidity sensor, an image collector and a radar.

[0033] Beneficial effects: During vehicle driving, the controller can assist the vehicle in adjusting driving operations such as speed, steering and braking according to information such as ambient temperature, humidity, road conditions and the driving status of the vehicle ahead to ensure driving safety.

[0034] Furthermore, a storage medium for target detection includes a storage device for storing image information, environmental information, and historical driving records of a vehicle in different driving environments.

[0035] Beneficial effects: The storage device can store image information, environmental information and historical driving records of the vehicle in different driving environments, which is convenient for the comprehensive analysis module to extract the required information for comprehensive analysis, and also convenient for the deep learning module to perform deep learning.

[0036] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The figure is a flow chart of a target detection optimization method in an embodiment of the present invention.

[0038] Figure 2 The figure is a schematic diagram of the steps of a target detection optimization method in an embodiment of the present invention.

[0039] Figure 3 It is an axonometric diagram of a target detection device in an embodiment of the present invention.

[0040] The figure numbers in the drawings of the specification include: 1. mounting bar; 2. temperature sensor; 3. humidity sensor; 4. image collector; 5. radar. DETAILED DESCRIPTION

[0041] The following is further described in detail through specific implementation methods:

[0042] Embodiment 1:

[0043] As attached Figure 1-Figure 3 As shown, a target detection optimization method includes the following steps:

[0044] S1, image acquisition: prepare the image acquisition module, which is used to acquire and analyze the image of the detection target; use the image acquisition module to acquire the image information around the vehicle, analyze the weather conditions and road conditions based on the image information, and select the pedestrians, vehicles, traffic facilities and obstacles in the image with different color frames; the image acquisition module identifies and distinguishes pedestrians, vehicles and traffic facilities based on their overall shape and contour features.

[0045] Road conditions include road type, road material, road markings and road bumps. Road types include straight sections, turning sections, uphill sections and downhill sections. Driving status includes vehicle speed, turning angle and lighting conditions. Obstacles include objects on the road other than pedestrians, vehicles, traffic facilities and road surfaces.

[0046] S2, environment acquisition: prepare the environment acquisition module, which is used to collect radar images, environmental temperature and environmental humidity to form thermal imaging and target cloud points; use the environment acquisition module to collect information about the vehicle's surrounding environment; the environment acquisition module collects the temperature of each detection target, displays the temperature image and temperature profile of the detection target in the form of thermal imaging, and compares and analyzes the temperature image and temperature profile of the detection target to determine pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment;

[0047] At the same time, the environment acquisition module is used to transmit electromagnetic wave signals from the vehicle's surrounding environment, receive echo signals, analyze and process the echo signals, extract the distance, size, shape, direction and speed of the detection target, judge the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment, and display them in the form of target cloud points;

[0048] The environment acquisition module will collect the environmental humidity and assist the image acquisition module in determining the current weather conditions.

[0049] The ambient temperature includes the ground surface temperature and the temperature inside the vehicle. The environmental acquisition module also collects the temperature inside the vehicle and adjusts the temperature inside the vehicle to the temperature value set by the driver.

[0050] S3, comprehensive analysis: prepare a comprehensive analysis module, which is used to receive and process the image information, thermal imaging and target cloud points collected by the image acquisition module and the environment acquisition module, and monitor the vehicle's driving status in real time; use the comprehensive analysis module to monitor the vehicle's driving status in real time, and combine the image information with the environmental information, and comprehensively analyze the pedestrians, vehicles, traffic facilities, road conditions and obstacles around the vehicle to obtain the current driving environment of the vehicle, and record the vehicle's driving environment and driving status each time to form a historical driving record;

[0051] S4, driving plan analysis: prepare a deep learning module, which is used to perform deep learning based on historical driving records and optimize driving plans; use the deep learning module to build a vehicle automatic driving model based on the convolutional layer algorithm according to the vehicle's historical driving records in different driving environments; simulate different driving plans based on the automatic driving model, combined with the current vehicle driving status and driving environment, and analyze the driving plan with a low accident rate and compliance with road traffic regulations based on the driving conditions simulated by different driving plans, the vehicle's driving status, the vehicle's surrounding environment and road traffic regulations.

[0052] The specific implementation process is as follows: first, a large amount of image recognition training is performed on the image acquisition module, and the image information around the vehicle is collected using the trained image acquisition module. During the image acquisition process, the image acquisition module will select and display pedestrians, vehicles, traffic facilities and obstacles in the image with different color frames, thereby reminding the driver of the pedestrians, vehicles, traffic facilities and obstacles that need to be paid attention to around his vehicle.

[0053] For example, there is a pedestrian crossing the zebra crossing in front of the driver's vehicle, the traffic light is red, and the remaining time is 20 seconds. There is a white car parked in the second lane of the opposite road; after the image recognition module recognizes the image, it will select the pedestrian with a red frame and move with the movement of the pedestrian in real time; the image recognition module will select the oncoming vehicle with a blue frame and move with the movement of the vehicle in real time; the image recognition module will select the traffic light with a green frame and highlight the remaining time of the red light of 20 seconds; at the same time, the image recognition module will also mark the lane outline, zebra crossing outline and the outlines of both sides of the road with white solid lines.

[0054] While collecting images, the environment collection module is used to collect the temperature conditions of each detection target. Since the temperatures of pedestrians, vehicles, traffic facilities, road surfaces and obstacles are all different, the temperature of vehicle engines is much higher than the body temperature of pedestrians, while the temperatures of road surfaces and obstacles are generally lower in normal weather. Therefore, the environment collection module uses thermal imaging to display the temperature image and temperature profile of the detection target. The lower the temperature, the bluer the image color, and the higher the temperature, the redder the image color. As the temperature rises, the image will change from dark blue to dark red. The environment collection module then compares and analyzes the temperature image and temperature profile of the detection target with the temperature images and temperature profiles of pedestrians, vehicles, traffic facilities, road surfaces and obstacles under normal circumstances, thereby effectively judging the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment.

[0055] For example, there is a pedestrian crossing the zebra crossing in front of the driver's vehicle, there is a right turn sign painted on the road surface of the first lane opposite, there is a car parked in the second lane of the road, and there is a puddle 10 meters in front of the vehicle; because the zebra crossing and traffic signs are made of different materials from the road surface, the puddle, manhole cover and drainage channel are different from the road surface structure and position, so their temperatures are also different. Therefore, thermal imaging will clearly show the pedestrians, vehicles, zebra crossings, right turn signs and puddles on the road surface with different temperature images and temperature contours.

[0056] Since the distance, speed, shape and size of different objects are different, the reception and reflection of waves will also be different. Since the propagation speed of electromagnetic waves is known, its speed is about the speed of light in a vacuum. The radar can calculate the distance between the target and the radar by measuring the time delay between the transmitted signal and the received echo. Specifically, the distance (R) is equal to the time delay (τ) multiplied by half of the electromagnetic wave propagation speed (c), that is, R = 2cτ; therefore, the environment acquisition module will also transmit electromagnetic wave signals from the vehicle's surrounding environment, receive echo signals, analyze and process the echo signals, extract the distance, size, shape, direction and speed of the detection target, and determine the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment; the environment acquisition module will determine the azimuth and elevation of the detection target relative to the radar through the joint scanning of multiple antennas, thereby determining the position of the detection target in three-dimensional space, and display the detection target in the form of target cloud points, so that the driver can understand the vehicle's surrounding environment more intuitively.

[0057] The environment acquisition module will collect the ambient humidity and transmit the ambient humidity data to the image acquisition module to assist the image acquisition module in determining the current weather conditions.

[0058] After the image acquisition and environment acquisition are completed, the comprehensive analysis module will extract the image acquisition and environment acquisition, give priority to judging the current weather conditions, and select the main judgment basis; if the weather conditions are hot, it is easy to cause the temperature difference between pedestrians, roads and obstacles to decrease during environment acquisition, making the temperature image in the thermal imaging similar in color and the temperature profile clarity reduced. At this time, the comprehensive analysis module will use the image information collected by the image acquisition module and the target cloud points collected by the radar as the main judgment basis to judge the pedestrians, vehicles, traffic facilities, road conditions and obstacles around the vehicle, and then use thermal imaging for auxiliary judgment; if it is raining, the image acquisition is prone to blurring, and the radar will also be affected. At this time, the vehicle engine temperature is still much higher than the pedestrian's body temperature, while the traffic facilities, road surface and obstacle temperatures will be lower than the pedestrian's body temperature. The comprehensive analysis module will use the temperature image and temperature profile collected by thermal imaging as the main judgment basis, and then use radar and image information for auxiliary judgment; in this way, the vehicle's surrounding environment can be detected under different weather conditions, achieving a more comprehensive detection effect.

[0059] The deep learning module will build a vehicle autonomous driving model based on the convolutional layer algorithm according to the vehicle's historical driving records in different driving environments; the autonomous driving model includes multiple convolutional layers, pooling layers and fully connected layers. The convolutional layer is used to extract features in the driving environment; the pooling layer is used to reduce the feature dimension and reduce the amount of calculation; the fully connected layer is used to map the extracted features to the final driving behavior to simulate different driving schemes, and analyze the driving conditions simulated according to different driving schemes based on the vehicle's driving status, the vehicle's surrounding environment and road traffic regulations. A driving plan with a low accident rate and compliance with road traffic regulations is analyzed.

[0060] The present invention uses the image acquisition module to recognize detailed images of the vehicle's surrounding environment, especially the real-time selection and display of pedestrians, vehicles, traffic facilities and obstacles, which significantly improves the driver's perception of the surrounding environment and reduces the risk of accidents caused by blind spots or negligence. At the same time, the environment acquisition module uses thermal imaging technology to effectively identify potential dangerous factors, such as puddles and obstacles on the road, even in complex or severe weather conditions, further improving driving safety.

[0061] The comprehensive analysis module can automatically adjust the judgment basis according to different weather conditions to ensure that the vehicle's surrounding environment can be effectively detected under various conditions. For example, in hot weather, when thermal imaging technology is limited by the reduction of temperature differences, the comprehensive analysis module will rely more on image information and radar data for judgment; on rainy days, when image and radar data are interfered with, the comprehensive analysis module will rely more on thermal imaging data. This flexible switching mechanism greatly enhances the environmental adaptability of the system.

[0062] The deep learning module builds an autonomous driving model by analyzing the vehicle's historical driving records in different driving environments. The deep learning module can simulate different driving plans, evaluate their accident risks and legality, and ultimately select the best driving plan. This decision-making method based on big data and machine learning enables vehicles to respond to complex traffic environments more intelligently and improve driving safety and efficiency.

[0063] Embodiment 2:

[0064] As attached Figure 1 , Figure 3 As shown, the difference from the above embodiment is that a target detection device includes a mounting bar 1, and the side wall of the mounting bar 1 is fixedly connected with a temperature sensor 2, a humidity sensor 3, an image collector 4 and a radar 5 by bolts.

[0065] The specific implementation process is as follows: the design of the mounting bar 1 can assemble the temperature sensor 2, the humidity sensor 3, the image collector 4 and the radar 5 into a whole, so as to better install the target detection device on the vehicle, thereby providing better target detection services for the vehicle.

[0066] Embodiment 3:

[0067] As attached Figure 1 , Figure 3 As shown, the difference from the above embodiment is that an electronic device for target detection includes a controller, and the controller is used to receive information collected by a temperature sensor, a humidity sensor, an image collector and a radar.

[0068] The specific implementation process is as follows: During vehicle driving, the controller can assist the vehicle in adjusting driving operations such as speed, steering and braking based on information such as ambient temperature, humidity, road conditions and the driving status of the vehicle in front to ensure driving safety.

[0069] Embodiment 4:

[0070] As attached Figure 1 , Figure 3 As shown, the difference from the above embodiment is that a storage medium for target detection includes a storage device, and the storage device is used to store image information, environmental information and historical driving records of the vehicle in different driving environments.

[0071] The specific implementation process is as follows: During vehicle driving, the storage device can store image information, environmental information and the vehicle's historical driving records in different driving environments, which is convenient for the comprehensive analysis module to extract the required information for comprehensive analysis, and also convenient for the deep learning module to perform deep learning.

[0072] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

Claims

1. A target detection optimization method, characterized in that: The following steps are involved: S1, image acquisition: prepare an image acquisition module, which is used to acquire and analyze images of the detection target; use the image acquisition module to acquire image information around the vehicle, analyze weather conditions and road conditions based on the image information, and select pedestrians, vehicles, traffic facilities and obstacles in the image with different color frames; S2, environment acquisition: prepare the environment acquisition module, which is used to collect radar images, environmental temperature and environmental humidity to form thermal imaging and target cloud points; use the environment acquisition module to collect the vehicle's surrounding environment information; The environment acquisition module collects the temperature of each detection target and displays the temperature image and temperature profile of the detection target in the form of thermal imaging. It compares and analyzes the temperature image and temperature profile of the detection target to determine the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment. At the same time, the environment acquisition module is used to transmit electromagnetic wave signals from the vehicle's surrounding environment, receive echo signals, analyze and process the echo signals, extract the distance, size, shape, direction and speed of the detection target, judge the pedestrians, vehicles, traffic facilities, road conditions and obstacles in the vehicle's surrounding environment, and display them in the form of target cloud points; The environment acquisition module will collect the environmental humidity and assist the image acquisition module in determining the current weather conditions; S3, comprehensive analysis: prepare a comprehensive analysis module, which is used to receive and process image information, thermal imaging and target cloud points collected by the image acquisition module and the environment acquisition module, and monitor the vehicle driving status in real time; The comprehensive analysis module is used to monitor the vehicle's driving status in real time, and the image information and environmental information are combined to comprehensively analyze the pedestrians, vehicles, traffic facilities, road conditions and obstacles around the vehicle to obtain the current driving environment of the vehicle, and record the driving environment and driving status of the vehicle each time to form a historical driving record; S4, driving plan analysis: prepare a deep learning module, which is used to perform deep learning based on historical driving records and optimize driving plans; A deep learning module is used to build a vehicle autonomous driving model based on the convolutional layer algorithm according to the vehicle's historical driving records in different driving environments. Different driving schemes are simulated based on the autonomous driving model and combined with the current vehicle driving status and driving environment. Driving conditions simulated based on different driving schemes are analyzed based on the vehicle driving status, the vehicle's surrounding environment and road traffic regulations to determine a driving plan with a low accident rate and compliance with road traffic regulations.

2. The target detection optimization method according to claim 1, characterized in that: In S1, the image acquisition module identifies and distinguishes pedestrians, vehicles and traffic facilities based on their overall shape and contour features.

3. The target detection optimization method according to claim 1, characterized in that: In S1, the road conditions include road type, road material, road markings and road unevenness.

4. The target detection optimization method according to claim 1, characterized in that: In S1, the road surface types include straight sections, turning sections, uphill sections and downhill sections.

5. The target detection optimization method according to claim 1, characterized in that: In S1, the driving status includes the vehicle's driving speed, turning angle, and light usage.

6. The target detection optimization method according to claim 1, characterized in that: In S1, obstacles include objects on the road other than pedestrians, vehicles, traffic facilities and road surfaces.

7. The target detection optimization method according to claim 6, characterized in that: In S2, the ambient temperature includes the ground surface temperature and the temperature inside the vehicle.

8. A target detection device based on a target detection optimization method according to any one of claims 1 to 7, characterized in that: It comprises a mounting bar (1), and a temperature sensor (2), a humidity sensor (3), an image collector (4) and a radar (5) are fixedly connected to the side wall of the mounting bar (1).

9. An electronic device for target detection based on a target detection optimization method according to any one of claims 1 to 7, characterized in that: The system comprises a controller, which is used for receiving information collected by a temperature sensor, a humidity sensor, an image collector and a radar.

10. A target detection storage medium based on a target detection optimization method according to any one of claims 1 to 7, characterized in that: It includes a storage device, which is used to store image information, environmental information and historical driving records of the vehicle in different driving environments.

Images