A Research of Game Cards Recognition System on Embedded Game Machine

• Main Authors: Bo-Shen Liu, 劉柏伸
• Other Authors: Chung-Ho Tsai
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/92113868201077029457

碩士 === 國立勤益科技大學 === 電子工程系 === 99 === Because of the influence by comic book and animation, playing the game cards has become a main stream for children. It is interesting to integrate game card identification function into the handheld embedded game console. Thus, this paper will investigate the image recognition scheme in identification application of the game cards for the handheld embedded game console. The whole skeleton, in this paper is divided into six parts including image extraction, color space conversion, edge detection, character segmentation, feature extraction and character recognition. First we capture the image of game cards’ number through the CMOS image sensor. Because the game cards are printed with many colors, we first transform the card image to YCbCr color space to achieve better effect with gray and binarization. After conducting edge detection, we then segment the characters that are to be recognized. Then the features are extracted from these binarized characters. Last the extracted features are compared with data in our database, and the game card could be identified as to be which one in this game.