Model based approach for effective segmentation of images based on background subtraction

Abstract

Image segmentation is considered as a vital part of image analysis which better understanding of these images is possible. Among the algorithms available to analyze images under motion, background subtraction is considered to be the most imperative. In this article an attempt is made to propose a methodology of image segmentation based on background subtraction by a proposing and developing a model based on truncated Gaussian distribution. The experimentation is carried on CDnet 2014 data set and results are analyzed using the metrics.