CAJ | 학술논문

为了更有效地分割手部图像，提出一种基于彩色阈值变换、形态学处理与模式识别的手部识别算法。利用手部与周围背景在RGB颜色分量上存在一定的差异，先通过RGB阈值变幻将彩色图像二值化，然后进行闭运算与孔洞填充，最后根据周长面积比去除背景中的孤立点集以提取手部。实验证明，该算法能够较为有效地将手部从背景中提取出来。
위료경유효지분할수부도상，제출일충기우채색역치변환、형태학처리여모식식별적수부식별산법。이용수부여주위배경재RGB안색분량상존재일정적차이，선통과RGB역치변환장채색도상이치화，연후진행폐운산여공동전충，최후근거주장면적비거제배경중적고립점집이제취수부。실험증명，해산법능구교위유효지장수부종배경중제취출래。
To segment hand image effectively,a hand recognition algorithm based on color threshold transform,morphologi?cal processing and pattern recognition is proposed in this paper. Based on the differences in RGB color components between hand and its background,the color image binaryzation is performed with RGB threshold change,the close operation and hole filling are carried out,and then hand image is extracted according to the perimeter?area ratio to remove isolated point sets. The experiment results show that the proposed algorithm can extract hand from the background effectively and accurately.