Foreground Background Separation and Tracking

摘要

Foreground background separation has become a significant factor in image processing for the past few years due to its many applications in the field of computer vision. In most researches, subtraction of images is used in the detection of foreground objects in an image. In this paper, the subtraction process will be substituted with a simple thresholding determined by genetic algorithm. Genetic algorithm solves for the ideal threshold that will be eventually used in classifying the pixels of an image is part of the foreground or the background. Erosion and dilation are other operations and processes used to come up with foreground objects. Tracking the foreground objects is another objective and will explore the capabilities of fuzzy logic most specifically the concept of fuzzy C-means (FCM) clustering.This study is part of a bigger research that aims to understand chicken’s behavioral patterns in a poultry set-up. The whole system has three main processes namely: genetic algorithm for determining the threshold, classification of pixels as part of object detection, and tracking of foreground objects.