Inherent Bias and Noise in the Hough Transform
IEEE Transactions on Pattern Analysis and Machine Intelligence
| Hi-index | 0.10 |
A method of image segmentation known as pyramid linking was developed several years ago for segmenting images into gray level subpopulations. This method is applied here to the segmentation of optical flow fields. The algorithm makes use of a hierarchical data structure, called a pyramid, in which an image has a stack of representations each of which has a coarser resolution than the one below. Each node constructs a model of the flow in the region of the image that it represents and then decides which of its fathers (on the level above) best represents this flow.