基于双目视觉的障碍物识别研究

Obstacle recognition based on binocular vision

随着各类无人技术的发展,对于障碍物检测技术的需求也日益提高。为实现快速精确的障碍物探测识别,文中基于双目立体视觉原理,利用张正友标定法对双目相机完成标定,获取内外参数,并对相机拍摄的图像进行去噪、去除亮度差及锐化处理,最终通过图像匹配得到目标场景的深度信息。根据应用需求对K-means聚类算法作出改进,以实现目标场景内障碍物的分割识别。实验结果表明,所设计的障碍物识别系统能够较好地实现障碍物的深度信息获取及分割识别,为后续避障工作提供了有效信息。

With the development of various types of unmanned technology,the requirements for obstacle detection technology are also increasing.In order to achieve fast and accurate obstacle detection and recognition,based on the principle of binocular stereo vision,this paper uses the ZHANG Zhengyou calibration method to complete the calibration of the binocular camera to obtain internal and external parameters,and then denoise,remove the brightness difference and sharpen the image taken by the camera,finally obtains the depth information of the target scene through image matching.The K⁃means clustering algorithm was improved and the obstacle recognition in the target scene was realized.The experimental results show that the designed obstacle recognition system can better achieve obstacle depth information acquisition and segmentation recognition,and provide effective information for the subsequent obstacle avoidance work.