Open AccessAngle and Scale Invariant Template Matching for Handling Image Distortions
Author(s) -
J Badrinaathan,
L. Srinivas
Publication year - 2018
Publication title -
international journal of engineering and technology
Language(s) - English
Resource type - Journals
ISSN - 2227-524X
DOI - 10.14419/ijet.v7i2.24.12009
Subject(s) - template matching , artificial intelligence , computer vision , pattern recognition (psychology) , computer science , histogram , invariant (physics) , template , histogram equalization , image (mathematics) , matching (statistics) , similarity (geometry) , adaptive histogram equalization , grayscale , mathematics , statistics , mathematical physics , programming language
Template matching is a diagnostic approach for detecting a patch of a template image in a given source image. This plays a vital role in multitudinal computer vision applications. In this paper, we propose a methodology that makes the naive template matching algorithm scale and angle invariant during the image recognition process where the source and template is converted to gray scale which makes the technique enhance its proficiency. The proposed algorithm handles the arbitrary modulations of the image patch with respect to size and angle by an exhaustive search of all combinations of sizes are done along with populous combinations of angles. The images adapted are subjected to certain filtering and convolution methods which deepens the quality of the images which in turn assists in retrieving the features with accuracy. The image intensities are adjusted using histogram equalization to enhance the image contrast. These images are then subjected to perform template matching using normalized cross correlation to measure similarity between those two images.