| Summary: | 碩士 === 南開科技大學 === 電子工程研究所 === 103 === This paper uses the Algorithm of Image processing to recognition poker. We try to use the Support Vector Machine and Match Template to recognition poker.
Experiment environment are fixed, like same distance to take a poker picture, and the same illuminant.
We also use simple Algorithm to help Image processing like transform the poker image to gray-level and change again to binary image. So, we can use the binary image to find contours. This contours parameter can use by set region of interest. However the region of interest can use the contours parameter to cut all of contours in the poker image, and use these parts to classify in the eighteen folders like A, number two to ten (zero), J, Q, K, four suits and other by using hand. After classify, use the Support Vector Machine to training and use Match Template to matching.
The experiment results, Support Vector Machine can through by train more examples to promote recognition rate. Match Template also have good recognition rate, but Match Template expend more time to matching
Keyword: Support Vector Machine, Match Template
|
|---|
