Fast and Accurate Template Matching Algorithm

Abstract

Template matching is a critical issue in the field of Artificial Intelligence (AI), with several applications including object identification and industrial inspection. The primary goal of this research was to analyse and propose a combined technique to improving the performance of template matching systems by combining image pyramid with the Sum of Absolute Difference (SAD) similarity metric. Based on the findings, the suggested method's capabilities may considerably increase the accuracy and execution speed of the template matching system. The experimental research relied on the utilisation of colour and grayscale pictures of varying sizes and lighting. The paper gave a full explanation of the experiments as well as an analysis of the proposed method's performance.