Target detection method based on monocular vision and millimeter wave radar fusion

Abstract

The invention relates to a target detection method based on monocular vision and millimeter-wave radar fusion. The method comprises the following steps: S1, acquiring a visual image and millimeter-wave radar data; s2, performing defogging preprocessing on the input visual image based on the average transmittance; s3, performing effective target screening according to the millimeter wave radar data; s4, fusing the effective target and the visual image based on coordinate transformation and time registration, and obtaining a region of interest in the fused visual image; s5, performing target identification on the region of interest by using a neural network, obtaining a visual identification result, and obtaining a millimeter wave radar identification result according to the millimeter waveradar data; and S6, performing weighted information decision on the millimeter-wave radar recognition result and the visual recognition result to obtain a final target detection result. Compared withthe prior art, the method has the advantages of improving the target recognition accuracy and reliability and the like.

Description

technical field [0001] The invention relates to the fields of computer vision and pattern recognition, in particular to a target detection method based on fusion of monocular vision and millimeter wave radar. Background technique [0002] Object detection is an essential part of the environment perception system for autonomous driving, and it is also a fundamental problem in computer vision. Road traffic target detection can use a variety of perception methods, such as cameras, millimeter-wave radars, and lidars. The camera image is processed, and the recognition of multiple types of traffic targets can be realized through artificial intelligence and deep learning technology; the distance and speed information of the target can be obtained according to the millimeter-wave radar echo signal and Kalman filtering technology; according to the point cloud characteristics of the lidar Target positioning and map reconstruction can be realized. These technologies have been widely ...

AI Technical Summary

Problems solved by technology

However, the single sensing source has defects and deficiencies due to the sensor itself, and the state information of the detection target is single. Therefore, target detection and recognition based on sensor information fusion is still the focus of research in the field of intelligent vehicle external environment perception.

However, in the existing technology, there is no effective fusion of millimeter-wave radar and visual information for target detection, and the existing fusion measures are only mechanically combined. There is no effective combination between the two, and the advantages of sensor fusion are not brought into play. Low accuracy and reliability of target detection results

Images