|Article title||PROBABILISTIC SEGMENTATION OF LOW-CONTRAST OPTICAL AND MOVING OBJECTS|
|Section||SECTION IV. MATHEMATICAL METHODS OF AN ARTIFICIAL INTELLECT|
|Month, Year||06, 2012 @en|
|Abstract||A new probabilistic method of improving low-contrast and small sized objects reliable detection on the images with hardly formalized background, under the conditions of full a priori uncertainty is considered. The notion of the Chernoff’s bound in the case of an object’s brightness and background arbitrary ratio basing on more peculiar Markov Random Field model and statistical approaches of classification and detection of objects is widened. The proposed dynamic segmentation allows forming binary fields with signs of objects’ motion on the gray or colour image sequence that for measuring parameters of motion in real time could be used.|
|Keywords||Segmentation; Markov random fields; Chernoff’s bound; Bayesian criterion; conditional probability; prior uncertainty.|
|References||1. Chernoff H. A measure of asymptotic efficiency for tests of a hypothesis based on the sum of observation // Annals of Mathematical Statistics. – 1952. – Vol. 23, № 4. – P. 493-507.
2. Lucas B.D., Kanade T. An Iterative Image Registration Technique with an Application to Stereo Vision // Proceedings of Imaging Understanding Workshop, Carnegie-Mellon University. – 1981. – P. 121-130.
3. Frucci M.; Sanniti di Baja G. From Segmentation to Binarization of Gray-level Images // Journal of Pattern Recognition Research. – 2008. – № 3 (1). – P. 1-13.