Criminal Identification Based on Androgenic Hair Pattern Using KNN-Clustering Method

Abstract

This paper is implemented in order to identify the criminals in forensics based on the androgenic hair pattern of the criminal particularly in Low-resolution images. The Proposed paper is more dependable than the existing forensic techniques to identify criminals. The paper is implemented in five stages; in the first stage the input images and the database images are obtained after which the database images are trained through the algorithm to get the trained images. In the second stage pre-processing of the images is done in order to convert the image to grey scale and remove noise using thresholding method. The Third stage the image is passed through a Gabor filter to detect the edges in the image, this is followed by the Fourth phase where KNN clustering method is used to obtain the region of interest followed by the application of the bilateral filter in order to enhance the image. In the final stage the indifferent value calculated from the input image is matched with all the indifferent values of the trained images to identify the criminal.