Fast and robust template matching algorithm in noisy image
Bong Gun Shin, So-Youn Park, Ju-Jang Lee · 2007
In the majority of robot applications, including human-computer interaction, template matching is used to find a specific area in a given image or a frame of video stream. Flexible and robust template matching algorithm necessitates feature extraction, for example gradient calculation. This requires complex calculation which causes bad response time of the system. An alternative solution is the use of index table, which stores coordinates that have the same grey level. However, due to the mechanism of the matching algorithm, it is necessary to have several disadvantages in the algorithm. But these restrictions are less important, and there is an idea that copes with these limitations. This paper proposes fast and robust template matching algorithm that uses grey level index table and image rank technique. This algorithm can find specific area under the given template query image with 30% Gaussian noise.