为克服Fylnn最小不连续相位展开算法效率较低的缺陷, 以快速高效地展开包裹相位图, 提出了一种基于边缘检测的最小不连续相位展开新算法。先对孤立噪声点预处理并将其排除在增长环搜寻扫描区域外, 然后利用边缘检测技术检测出相位不连续区域, 并把其作为扫描区域。由于利用边缘检测使增长环搜寻所需的扫描范围在很大程度上缩小, 最小不连续相位展开算法的效率得到了很大提高。详细描述了新算法的原理和实现步骤, 并用计算机模拟和实际的包裹相位图来验证其有效性。相位展开结果表明, 新算法能正确展开这些包裹相位图, 且较最小不连续算法有更高的效率。

A new minimum-discontinuity unwrapping algorithm based on edge detection is proposed to overcome the disadvantage of Flynn′s minimum-discontinuity approach and unwrap the wrapped phase images rapidly and efficiently. In the new algorithm, isolated noisy pixels are excluded from the scanning areas of searching for improving loops firstly, and then the phase′s discontinuous areas within wrapped phase image are detected based on edge detection techniques and taken as the scanning areas. Because the required scanning range is cut down to a large extent by use of edge detection methods, the efficiency of the minimum-discontinuity phase unwrapping algorithm is significantly improved. The principle and implementation steps of the new algorithm are described in detail, and some computer simulation and experimental wrapped phase images are used to verify the effectiveness of the algorithm. Unwrapped results show that the proposed algorithm works perfectly and is more effective than the Flynn′s minimum discontinuity-approach.