A novel automatic thresholding segmentation method with local adaptive thresholds

TL;DR

This paper introduces a new automatic segmentation method for grayscale images that uses local adaptive thresholds based on gradient analysis, effectively handling fuzzy boundaries and mimicking human perception.

Contribution

It presents a novel thresholding segmentation technique that adaptively determines local thresholds by analyzing boundary gradients, improving segmentation accuracy for fuzzy object boundaries.

Findings

Effective segmentation of bright objects from dark backgrounds

Handles fuzzy boundaries well

Highly automatic and mimics human perception

Abstract

A novel method for segmenting bright objects from dark background for grayscale image is proposed. The concept of this method can be stated simply as: to pick out the local-thinnest bands on the grayscale grade-map. It turns out to be a threshold-based method with local adaptive thresholds, where each local threshold is determined by requiring the average normal-direction gradient on the object boundary to be local minimal. The method is highly automatic and the segmentation mimics a man's natural expectation even the object boundaries are fuzzy.