A Markov Random Field and Active Contour Image Segmentation Model for Animal Spots Patterns

摘要

Non-intrusive biometrics of animals using images allows to analyze phenotypicpopulations and individuals with patterns like stripes and spots withoutaffecting the studied subjects. However, non-intrusive biometrics demand a welltrained subject or the development of computer vision algorithms that ease theidentification task. In this work, an analysis of classic segmentationapproaches that require a supervised tuning of their parameters such asthreshold, adaptive threshold, histogram equalization, and saturationcorrection is presented. In contrast, a general unsupervised algorithm usingMarkov Random Fields (MRF) for segmentation of spots patterns is proposed.Active contours are used to boost results using MRF output as seeds. As studysubject the Diploglossus millepunctatus lizard is used. The proposed methodachieved a maximum efficiency of $91.11\%$.