Implementing and Optimizing Template Matching Techniques for Home Automation

Background/Objectives: In analyzing the modern world with huge amounts of digital visuals, the image analysis techniques play a major role. It helps in monitoring the organizing the visual data. The contents or objects in the image have significance on image recognition. Methods/Statistical Analysis: There are various algorithms which recognize the real world object. This system proposes a model which uses modified template matching technique based on SURF algorithm and squared difference error method. Findings: The template matching is done based on image features comparison. SURF algorithm of template matching is based on keypoint extraction from images whereas the squared difference error algorithm of template matching is based on pixel feature comparison. To analyze the performance of the algorithms in real world, both the algorithms been implemented using embedded microcontroller Raspberry pi and Pi Camera. Conclusion: The Electrical appliances used in home were considered as objects to be recognized. The appliances the detected via Pi Camera. This algorithm can be further explored in areas like multiple object recognition and various other autonomous machine vision system.