基于双目的目标快速测距算法
FAST TARGET RANGING ALGORITHM BASED ON BINOCULAR VISIONS
李学森 1李自立1
作者信息
- 1. 广西师范大学电子工程学院 广西桂林 541000
- 折叠
摘要
为保证测距精度的同时提高计算效率,针对特定的应用场景,提出一种在Census变换理论基础上改进的立体匹配算法.以降低计算的像素需求,优化测距方法为途径,针对双摄像头获取的视频图像,通过图像灰度化、像素值平均、像素定标比较以及像素点替换等图像分析方法,实现在特定应用场景下提高双目摄像头的数据计算效率,加快硬件的运算速度,达到算法快速、稳定、有效地运行的目的.根据实验结果分析,该算法在图像噪声抑制上具有较好的效果,在实际应用中展现了较好的稳定性与较高的计算效率.
Abstract
In order to improve the computational efficiency and ensure the accuracy of ranging,a stereo matching algorithm based on improved Census transformation theory is proposed for specific application scenarios.In order to reduce the pixel requirements required for calculation,the optimized ranging method was adopted to improve the data computing efficiency of binocular cameras in specific application scenarios and speed up the computing speed of hardware through image analysis methods such as image graying,pixel value averaging,pixel calibration comparison and pixel point replacement for video images obtained by dual cameras.The fast,stable and effective operation of the algorithm was achieved.According to the analysis of experimental results,the algorithm has a good effect on image noise suppression,and shows better stability and higher computational efficiency in practical application.
关键词
双目视觉/测距/改进Census变换/目标视差区域Key words
Binocular vision/Distance measurement/Improved Census transformation/Target parallax region引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
维普
万方数据