摘要
立体匹配是计算机视觉中的一项重要技术。文中对立体匹配中引入极线约束的SIFT算法和经典SIFT算法进行了比较,通过仿真实验证明,在处理尺度缩放、亮度变化的图像时,引入极线之后的SIFT算法的立体匹配效果明显提升。根据三维成像原理,对SIFT算法提取的特征点进行三维重建,获得了目标场景的视差图,实现了测距。
Stereo matching is an important technique in computer vision.This paper compares the SIFT algorithm and the classical SIFT algorithm with the introduction of the polar line constraint in stereo matching.The simulation results show that when the image is scaled and the brightness changes,the stereo matching effect of the SIFT algorithm after the introduction of the pole line is significantly improved.According to the principle of 3Dimaging,the feature points extracted by SIFT algorithm are reconstructed,and the disparity map of the target scene is obtained.
出处
《通信电源技术》
2016年第2期66-67,共2页
Telecom Power Technology
关键词
双目测距
SIFT
立体匹配
binocular distance measurement
SIFT
stereo matching
