A new threshold selection method for separation of objects and background

Abstract

A new threshold selection algorithm for objects and background separation using a histogram of the number of foreground regions was proposed. This method used a human separation method for objects and background, examined the changes in the number of foreground regions in the image with varying threshold and determined a proper threshold value. The proposed technique was applied to several images and was compared with other histogram-based segmentation methods. The possibility of extension to multi-level thresholding was also examined.