摘要
通过分析目前立体视差估计方法的不足,提出一种基于冗余小波变换的视差估计新算法。首先对参考图像进行冗余小波变换,提取特征点,形成DT(Delaunay Triangle)网格,然后根据三角形特征点在目标图像中进行视差估计,最后通过仿射变换形成浓密视差图。实验表明该算法能有效获得视差矢量,视差匹配后能得到良好的重建图像。
By analysing the shortage of the current stereo disparity estimation methods, a new method based on redundant wavelet transform is proposed. First, The redundant wavelet transform is used to obtain the characteristic points of reference picture and generate a DT（Delaunay Triangle） mesh. And then, according to the characteristic points of triangles, the dis parity estimation is implemented in target image. Finally, the dense disparity map is generated by using affine transformation, experimental results show that the new method can obtain the disparity vectors and the well reconstruction of original image.
出处
《电视技术》
北大核心
2008年第4期23-25,共3页
Video Engineering
关键词
视差估计
冗余小波
视差匹配
立体图像
disparity estimation
redundant wavelet
disparity matching
stereo image
