Image matching algorithm based on grayscale and its improvement
Kun Jia · 2013
Image matching algorithm is composed of two parts, which are the algorithm based on the grayscale of the image directly and the algorithm based on the feature of physical shape and the advanced feature of the image. In this paper, two algorithms (correlation coefficient and histogram) based on the grayscale were analyzed and improved. During the process of the algorithm of histogram, two sequences (referential sequence and comparative sequence) are created according to the grayscale of the template image and the real-time sub-image. Compare and calculate the relational degree between the two sequences. The higher the relational degree is the closer to the best matching position. And the algorithm of histogram was improved through the cosine correlation to increase the matching velocity. Both of the two algorithms use the principles (from coarsely to finely, from partial template to whole, interlaced every column) to achieve improvement.