Quantization and equalization of pseudocolor images in hand thermography

Abstract

Thermal images consist of a single matrix, though several colorization algorithms are mostly applied to highlight the differences in radiation. The colorization is done by built-in colormaps which usually conceal some features, so in case of being already devoted to a specific purpose, hardly useful for other requirements. We therefore proposed a novel quantization and indexing algorithm in hand thermography. The images already extracted by pseudocolor map are segmented by red-channel conversion for turning into grayscale and 4-level quantization. Moreover, through the histograms and cumulative histograms, intermediary indexing is applied by Midway image equalization. The results seem promising for revealing the details with high contrast differences in quantization.