A robust algorithm for segmenting deformable linear objects fromvideo image sequences

摘要

A new algorithm for segmenting and tracking deformable(shape-changing) linear (thin and long with negligible diameter) objectsin video images is presented. The core of algorithm is to detect the twoboundary curves of a deformable linear object within adaptive trackingwindows. The boundaries are found by analyzing the gray value gradientswithin the tracking windows. The algorithm allows moderate changes inthe image brightness by introducing tolerance parameters to thealgorithm. It also provides a robust tracking of the shape- andwidth-changing object by using a fuzzy estimation of the object'smembership function at the tracked points. The result describes theshape of the object and is used to define features for shape changedetection