Environment sensing method and system for autonomous vehicle

Abstract

The invention relates to the technical field of automatic driving, in particular to an environment sensing method and system for an automatic driving vehicle, and solves the problems that in the prior art, the sensing ability of a sensor may be seriously influenced under the severe weather condition, the sensing ability of a camera may be greatly reduced under the light condition, at night or under the insufficient light condition, and the sensing ability of the camera cannot be greatly influenced. And the surrounding environment cannot be accurately identified. The invention discloses an environment sensing method and system for an autonomous vehicle. The system comprises data acquisition, data preprocessing, data fusion, target detection and tracking, decision and planning, a camera and a laser radar. According to the invention, through the cameras distributed at different positions of the vehicle, an omnibearing visual field can be captured, it is ensured that the vehicle comprehensively understands the surrounding environment, the 360-degree rotation function of the laser radar enables the vehicle to obtain accurate three-dimensional data, important information is provided for subsequent navigation and obstacle avoidance, and the millimeter wave radar and the ultrasonic sensor are used.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to an environment perception method and system for an autonomous driving vehicle. Background Art

[0002] The brain system realizes the unmanned intelligent car. The self-driving vehicle relies on artificial intelligence, visual computing, radar, monitoring devices and other technologies to achieve safe driving without any human active operation. These technologies together constitute the core system of the self-driving vehicle, enabling it to perceive the surrounding environment, make decisions and perform corresponding driving actions. Early research and development: The concept and technology research and development of self-driving cars have a history of decades. Practical trend: In the early 21st century, self-driving technology began to show a trend of approaching practical application. Sensor technology: Self-driving vehicles are usually equipped with a variety of sensors, such as high-definition cameras, lidar, etc., which are used to capture and perceive information about the surrounding environment. Artificial intelligence technology: Through advanced algorithms and machine learning technology, self-driving vehicles can identify and process data captured by sensors to make driving decisions. Global Positioning System: GPS and other global positioning system technologies provide precise positioning information for self-driving vehicles to ensure that they travel along the predetermined route. Road testing: Many places around the world have conducted road tests of self-driving vehicles to verify their safety and feasibility. Commercial application: Some self-driving technologies have begun to be commercialized, such as self-driving taxi services.

[0003] Environmental perception of autonomous vehicles is one of its core technologies. It involves how the vehicle understands and responds to the surrounding road and traffic conditions. Environmental perception refers to the ability of autonomous vehicles to obtain and analyze information about the surrounding environment through various sensors. This is the basis for realizing autonomous driving, because only by fully perceiving and understanding the environment can the vehicle make correct driving decisions. The environmental perception of autonomous vehicles mainly relies on a variety of sensors. Environmental perception is not just about simply collecting sensor data, but also about fusing and processing this data. Through the fusion of data from multiple sensors, autonomous vehicles can obtain more comprehensive and accurate environmental information. The data fusion algorithm will comprehensively consider the advantages and limitations of different sensors, thereby improving the reliability and robustness of the perception system.

[0004] Although the environmental perception technology of autonomous vehicles in the existing technology has made significant progress, it still has some shortcomings. Weather impact: Under severe weather conditions, the perception ability of the sensor may be seriously affected. Light conditions: At night or in low light conditions, the perception ability of the camera will drop significantly, which may lead to the inability to accurately identify the surrounding environment. Limited perception range: For example, the detection angle of millimeter-wave radar is relatively small, and multiple radars are required to cover a wider range. Resolution problem: The resolution of some sensors may not be high enough, resulting in the inability to accurately identify distant objects or details. Data fusion difficulty: Data from different sensors need to be effectively fused and processed to obtain accurate environmental information. However, data fusion algorithms may face challenges in complexity and real-time performance. Computing resource requirements: A large amount of sensor data requires powerful computing resources for real-time processing and analysis, which places higher requirements on the vehicle's hardware system. Summary of the invention

[0005] The purpose of the present invention is to provide an environmental perception method and system for an autonomous driving vehicle, which solves the problem in the prior art that the perception ability of the sensor may be seriously affected under severe weather conditions and light conditions. At night or in low light conditions, the perception ability of the camera will be greatly reduced, which may lead to the inability to accurately identify the surrounding environment.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] An environmental perception method and system for an autonomous driving vehicle, including data collection, data preprocessing, data fusion, target detection and tracking, decision-making and planning, using multiple devices such as cameras, lidar, millimeter-wave radar and ultrasonic sensors to comprehensively collect visual and distance information around the vehicle, clean and optimize the original sensor data, eliminate noise and interference, ensure data accuracy and consistency, prepare for subsequent processing, combine data from different sensors, integrate through algorithms, build a comprehensive and accurate model of the vehicle's surrounding environment, identify and continuously track key targets in the surrounding environment, such as other vehicles, pedestrians, etc., ensure that the autonomous driving system understands the surrounding dynamics, formulates safe and effective driving strategies based on the perceived environmental information, and plans the best driving path.

[0008] Preferably, the visible light images are captured by cameras to obtain visual information such as roads, vehicles, pedestrians, etc. These cameras may be distributed in different positions of the vehicle to provide a full range of vision. The laser radar is used to emit laser beams and measure the distance, shape and position of objects by receiving the reflected lasers. The laser radar can rotate 360 ​​degrees to obtain three-dimensional data of the surrounding environment. Millimeter wave radar and ultrasonic sensors detect objects by emitting and receiving radio waves or sound waves and provide information such as distance and speed. These sensors are usually installed on the front and rear bumpers or sides of the vehicle for close-range obstacle detection.

[0009] Preferably, the collected raw data is preprocessed, including operations such as denoising, filtering and enhancement, to improve the quality and availability of the data, and the data of different sensors are time synchronized and spatially aligned to ensure that the information provided by each sensor can match and fuse with each other.

[0010] Preferably, a data fusion algorithm is used to integrate data from different sensors to obtain more comprehensive and accurate environmental information. A weighted fusion and Kalman filtering method are used to combine data from multiple sensors to model the surrounding environment and generate real-time environmental information required for vehicle driving.

[0011] Preferably, the fused data is processed by computer vision and machine learning algorithms to identify and track obstacles, vehicles, pedestrians and other targets on the road, and to classify and locate the detected targets so that the vehicle can respond appropriately.

[0012] Preferably, based on the perceived environmental information, the decision-making system of the autonomous driving vehicle will formulate a driving strategy according to preset rules and algorithms, plan a safe driving path, and take into account factors such as road conditions, traffic regulations, and the behavior of other vehicles.

[0013] The present invention has at least the following beneficial effects:

[0014] The present invention can capture a full range of vision through cameras distributed at different positions of the vehicle, ensuring that the vehicle has a comprehensive understanding of the surrounding environment. The 360-degree rotation function of the laser radar enables the vehicle to obtain accurate three-dimensional data, providing important information for subsequent navigation and obstacle avoidance. The use of millimeter-wave radar and ultrasonic sensors enables the vehicle to accurately detect obstacles at close range, improving driving safety. Through pre-processing operations such as denoising and filtering, the quality of raw data is significantly improved, laying the foundation for subsequent data processing and analysis, ensuring that the information provided by different sensors can match and fuse with each other, thereby improving the accuracy and reliability of the perception system. Through the data fusion algorithm, the data from different sensors are integrated to obtain more comprehensive and accurate environmental information. It provides strong support for the decision-making of autonomous vehicles. It combines data from multiple sensors to model the surrounding environment in real time, allowing the vehicle to perceive and understand the external environment more accurately. Through computer vision and machine learning algorithms, it can accurately identify and track obstacles, vehicles and pedestrians on the road, provide real-time navigation information for autonomous vehicles, classify and locate detected targets, help vehicles make correct driving decisions, and improve driving safety. Based on comprehensive and accurate environmental information, the decision-making system of autonomous vehicles can formulate more reasonable and safe driving strategies, taking into account factors such as road conditions, traffic rules and the behavior of other vehicles, and plan safe driving routes to ensure that autonomous vehicles can drive safely and efficiently in various complex environments. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0016] Embodiment 1

[0017] It includes data collection, data preprocessing, data fusion, target detection and tracking, decision-making and planning, and uses a variety of equipment such as cameras, lidar, millimeter-wave radar and ultrasonic sensors to collect visual and distance information around the vehicle in all directions, clean and optimize the original sensor data, eliminate noise and interference, ensure data accuracy and consistency, and prepare for subsequent processing. It combines data from different sensors and integrates them through algorithms to build a comprehensive and accurate model of the vehicle's surrounding environment, identify and continuously track key targets in the surrounding environment, such as other vehicles and pedestrians, to ensure that the autonomous driving system understands the surrounding dynamics, formulates safe and effective driving strategies based on the perceived environmental information, and plans the best driving path.

[0018] Cameras distributed at different positions on the vehicle can capture a full range of views, ensuring the vehicle's comprehensive understanding of the surrounding environment. The 360-degree rotation function of the lidar enables the vehicle to obtain accurate three-dimensional data, providing important information for subsequent navigation and obstacle avoidance. The use of millimeter-wave radar and ultrasonic sensors enables the vehicle to accurately detect obstacles at close range and improve driving safety. Through pre-processing operations such as denoising and filtering, the quality of the original data is significantly improved, laying the foundation for subsequent data processing and analysis, ensuring that the information provided by different sensors can match and fuse with each other, thereby improving the accuracy and reliability of the perception system.

[0019] Embodiment 2

[0020] Visible light images are captured by cameras to obtain visual information such as roads, vehicles, and pedestrians. These cameras may be distributed in different positions of the vehicle to provide a full range of vision. LiDAR emits laser beams and measures the distance, shape, and position of objects by receiving the reflected lasers. LiDAR can rotate 360 ​​degrees to obtain three-dimensional data of the surrounding environment. Millimeter-wave radar and ultrasonic sensors detect objects by emitting and receiving radio waves or sound waves and provide information such as distance and speed. These sensors are usually installed on the front and rear bumpers or sides of the vehicle for close-range obstacle detection.

[0021] Embodiment 3

[0022] The collected raw data is preprocessed, including denoising, filtering, and enhancement operations, to improve the quality and availability of the data. The data from different sensors are synchronized in time and aligned in space to ensure that the information provided by each sensor can be matched and fused with each other.

[0023] Embodiment 4

[0024] The data fusion algorithm is used to integrate data from different sensors to obtain more comprehensive and accurate environmental information. The weighted fusion and Kalman filtering methods are used to combine the data from multiple sensors to model the surrounding environment and generate real-time environmental information required for vehicle driving.

[0025] Embodiment 5

[0026] The fused data is processed through computer vision and machine learning algorithms to identify and track obstacles, vehicles, pedestrians and other targets on the road, and the detected targets are classified and located so that the vehicle can respond appropriately.

[0027] Embodiment 6

[0028] Based on the perceived environmental information, the decision-making system of the autonomous driving vehicle will formulate driving strategies according to preset rules and algorithms, plan a safe driving path, and take into account factors such as road conditions, traffic rules, and the behavior of other vehicles.

[0029] Through data fusion algorithms, data from different sensors are integrated to obtain more comprehensive and accurate environmental information, providing strong support for the decision-making of autonomous vehicles. By combining data from multiple sensors, the surrounding environment is modeled in real time, enabling the vehicle to perceive and understand the external environment more accurately. Through computer vision and machine learning algorithms, obstacles, vehicles and pedestrians on the road can be accurately identified and tracked, providing real-time navigation information for autonomous vehicles, classifying and locating detected targets, helping vehicles make correct driving decisions and improving driving safety. Based on comprehensive and accurate environmental information, the decision-making system of autonomous vehicles can formulate more reasonable and safe driving strategies, taking into account factors such as road conditions, traffic rules and the behavior of other vehicles, and planning safe driving routes to ensure that autonomous vehicles can drive safely and efficiently in various complex environments.

[0030] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. An environmental perception method and system for an autonomous driving vehicle, including data collection, data preprocessing, data fusion, target detection and tracking, decision-making and planning, cameras, lidar, millimeter-wave radar, ultrasonic sensors and other equipment, to collect visual and distance information around the vehicle in all directions, clean and optimize the original sensor data, eliminate noise and interference, ensure data accuracy and consistency, prepare for subsequent processing, combine data from different sensors, integrate through algorithms, build a comprehensive and accurate model of the vehicle's surrounding environment, identify and continuously track key targets in the surrounding environment, such as other vehicles, pedestrians, etc., to ensure that the autonomous driving system understands the surrounding dynamics, formulates safe and effective driving strategies based on the perceived environmental information, and plans the best driving path.

2. The method and system for environmental perception of an autonomous driving vehicle according to claim 1, characterized in that: The cameras are used to capture visible light images and obtain visual information such as roads, vehicles, and pedestrians. These cameras may be distributed in different positions of the vehicle to provide a full range of vision. LiDAR is used to emit laser beams and measure the distance, shape, and position of objects by receiving the reflected lasers. LiDAR can rotate 360 ​​degrees to obtain three-dimensional data of the surrounding environment. Millimeter-wave radar and ultrasonic sensors detect objects by emitting and receiving radio waves or sound waves and provide information such as distance and speed. These sensors are usually installed on the front and rear bumpers or sides of the vehicle for close-range obstacle detection.

3. The method and system for environmental perception of an autonomous driving vehicle according to claim 1, characterized in that: The collected raw data is preprocessed, including operations such as denoising, filtering and enhancement, to improve the quality and availability of the data, and the data of different sensors are synchronized in time and aligned in space to ensure that the information provided by each sensor can match and fuse with each other.

4. The method and system for environmental perception of an autonomous driving vehicle according to claim 1, characterized in that: The data fusion algorithm is used to integrate data from different sensors to obtain more comprehensive and accurate environmental information. The weighted fusion and Kalman filtering methods are used to combine the data from multiple sensors to model the surrounding environment and generate real-time environmental information required for vehicle driving.

5. The method and system for environmental perception of an autonomous driving vehicle according to claim 1, characterized in that: The fused data is processed through computer vision and machine learning algorithms to identify and track obstacles, vehicles, pedestrians and other targets on the road, and the detected targets are classified and located so that the vehicle can respond appropriately.

6. The method and system for environmental perception of an autonomous driving vehicle according to claim 1, characterized in that: Based on the perceived environmental information, the decision-making system of the autonomous driving vehicle will formulate a driving strategy according to preset rules and algorithms, plan a safe driving path, and take into account factors such as road conditions, traffic regulations, and the behavior of other vehicles.