基于整数图像技术的自适应窗口立体匹配算法

针对固定窗口区域匹配算法在深度不连续处容易产生相对较大匹配误差的问题,提出了一种基于整数图像技术的自适应窗口立体匹配新算法.文中一种新的相似性测度函数和四方向线型掩模的采用不仅减少了噪声影响而且提高了匹配准确度.同时利用由窗口内平均误差、误差方差以及较大窗口偏移组成的窗口代价函数来确定窗口尺寸极大地提高了匹配精度;整数图像技术的引入将匹配时运算量降低为一常数.实验表明基于整数图像技术的自适应窗口立体匹配算法在局部匹配方法中是比较优秀的,便于实时实现.

Adaptive Window for Stereo Matching Algorithm using Integral Image

Because the fixed window area - based stereo algorithms always give rise to relatively large matching errors around depth discontinuities, an adaptive window for stereo matching algorithm using integral image was proposed. In this paper, a new similarity function and Four - direction line mask were employed not only to reduce noise effects but also to improve matching accuracy. Meanwhile, the window cost function that consists of average errors, variance of error and bias to larger windows was used to determine the size of matching and improve the matching accuracy. Using integral image technology while computing matching cost takes constant time. At last, the experiment result shows the advantages of this algorithm, and it can be used in real - time system.