Improving Detection Results of Differential Background Subtraction Method by Merging it with C-HDC Algorithm

Abstract

A method which allows to improve object detection abilities of conventional difference-based background subtraction algorithm has been presented in this paper. The algorithm is a hybrid method consisting of the background subtraction method and the Contour Histogram Displacement Calculation (C-HDC) algorithm. Background subtraction method processes a greyscale image, while C-HDC calculates displacement values basing on its contour histograms. Tracking ratios before, and after merging with C-HDC have been measured and presented.