Quick identification method for object vehicle lane changing

Abstract

A method for quickly identifying lanes-change condition of object vehicle includes utilizing machine vision and radar to collect horizontal and longitudinal distances, relative speed and the same lane possibility between object vehicle and local vehicle, making comparison and quantization on said data collected each time, making compensation and tracking prediction on relevant data, calculating confidence of each vehicle by using two layers of neural network training to obtain object vehicle weight and setting object vehicle to be dangerous object once calculated confidence coefficient is over confidence threshold.

Description

Technical field: [0001] The invention relates to a rapid identification method under the lane-changing condition of a target vehicle, which belongs to the technical field of intelligent vehicle driving environment perception. Background technique: [0002] In the research field of longitudinal active safety, selecting dangerous targets or car-following targets from the identified candidate vehicles is the basis for subsequent collision avoidance warning or automatic driving. Traditional smart vehicle longitudinal active safety systems mostly use radar as the sensor for target vehicle identification, but the false alarm rate of radar is high, and structures on the side of the road and above the road may be misdetected by radar as obstacles in the driving area in front of the vehicle And its detection field of view is limited. When the road has a large curvature, the vehicle in front is easy to drive out of the radar's measurement range. In addition, when the target vehicle ch...

AI Technical Summary

Problems solved by technology

The above two methods do not detect lane lines, and it is difficult to make an accurate judgment on the relative positional relationship between the target vehicle and the lane

Document 3 (CladyX, et al. Object tracking with a Pan-tilt-zoom camera: Application to car driving assistance, 2001) uses an ordinary camera to identify the lane lines and driving vehicles ahead, and uses a focal length, angle The variable camera (Pan-Tilt-Zoom Camera) tracks long-distance targets, but the system needs to coordinate the calibration of the two cameras, and the equipment composition and algorithm are very complicated

Document 4 (CHU Jiangwei, et al. Study on method of detecting preceding vehicle based on monocular camera, 2004) first detects the lane line, and then uses the two detected straight lines to form a triangular area, in which the target vehicle is recognized , the problem is that it is not easy to respond in time to vehicles cutting into the lane

[0004] The above method does not fully integrate and utilize the horizontal and vertical perception information, which leads to insufficient stability of target vehicle identification and tracking, and when the target vehicle changes lanes, the algorithm does not respond to the state change of the target vehicle in a timely manner.

Images