The invention discloses an automatic driving risk pre-judgment system and method based on improved YOLOV7-Tiny and SS-LSTM in a V2X environment, and the system comprises a visual recognition and tracking module which is used for photographing an image around a vehicle, and carrying out the target detection of the photographed visual image; the speed and distance measuring module is used for classifying the detected target vehicle into a social vehicle and an intelligent network connection vehicle to obtain the position and speed of the social vehicle and the distance between the social vehicle and the vehicle; the trajectory prediction module is used for continuously correcting the long-term prediction in the LSTM algorithm through the short-term prediction of Kalman filtering in the StrongSORT algorithm so as to establish an SS-LSTM trajectory prediction model, and the distance between the two vehicles and the position and speed of the target vehicle are used as the input of the model to obtain the driving trajectory of the target vehicle; the vehicle networking communication module is used for obtaining the speed and position of the intelligent networking vehicle and the distance between the two vehicles; and the risk pre-judgment module is used for judging whether collision risks exist between the target and the own vehicle according to the driving tracks of the target and the own vehicle. The accuracy of automatic driving risk pre-judgment is improved.
一种V2X环境下基于改进YOLOV7‑Tiny和SS‑LSTM的自动驾驶风险预判系统和方法,包括:视觉识别跟踪模块,用于对自身车辆周围的图像进行拍摄,对拍摄得到的视觉图像进行目标检测;测速测距模块,用于将检测到的目标车辆分类为社会车辆和智能网联车辆,得到社会车辆的位置、速度及与自身车辆的两车距离;轨迹预测模块,用于通过StrongSORT算法中卡尔曼滤波的短期预测来不断矫正LSTM算法中的长期预测,以建立SS‑LSTM轨迹预测模型,将两车距离和目标车辆的位置、速度作为模型的输入得到目标车辆的行驶轨迹;车联网通信模块,用于获得智能网联车辆的速度、位置及两车距离;风险预判模块,用于根据目标和自身车辆的行驶轨迹判断两者是否存在碰撞风险。其提高了自动驾驶风险预判的准确性。
Automatic driving risk pre-judgment system and method based on improved YOLOV7-Tiny and SS-LSTM in V2X environment
V2X环境下基于改进YOLOV7-Tiny和SS-LSTM的自动驾驶风险预判系统及方法
2024-01-19
Patent
Elektronische Ressource
Chinesisch
Rapid visual detection method of highway pavement sag based on improved Yolov7-tiny
DOAJ | 2024
|Freeway vehicle detection on improved YOLOv7
SPIE | 2023
|