On the application of the enhanced Hoshen-Kopelman algorithm for image analysis

The enhanced Hoshen-Kopelman (EHK) algorithm is introduced for a single pass analysis of spatial clusters in large images. The EHK algorithm is a generalisation of the algorithm known in the statistical physics literature as the Hoshen-Kopelman (HK) algorithm. While the HK algorithm was designed to compute cluster sizes in a binary image, the EHK algorithm enables the computation of cluster shape parameters such as spatial clusters moments, perimeters and bounding boxes in a multiple class image. An example of spatial cluster analysis for a simulated image containing 2 x 10(9) pixels is given. (C) 1998 Elsevier Science B.V. All rights reserved.