摘要
在羽毛球生产中,有缺陷的羽毛必须分拣出来,目前主要依靠人工完成。机器视觉可以被应用于羽毛自动分拣邻域以提高生产效率。文中针对羽毛缺陷的阶梯形状修剪问题,提出改进的弯曲度算法,对羽毛边缘曲线的拐点进行定位。在原有算法的基础上,增加了对拐点附近羽毛边缘形状的判断。考虑阶梯修剪的特有模式,通过模式识别的方式对羽毛阶梯修剪特征进行识别。提高了算法的稳定性,降低误判率。可区分阶梯修剪和分叉等其它缺陷。通过实际检验,该算法判断的结果是可靠的。
In the production of badminton, the feather which has disfigurement must be collected. Now, it is mainly depended on human. Machine vision could be used in this area to improve the efficiency. To solve the problem of the step form disfigurement of the feathers, an improved algorithm based on the tortuosity was presented here to detect comers on contour in digital image. Focusing on the specific mode of step form disfigurement, the algorithm increased the stability and decreased the error rate. By using this algorithm the step form and furcation disfigurement or other similar ones can be differentiated correctly. The practical production proved that the algorithm is credible.
出处
《计算机技术与发展》
2012年第3期166-168,172,共4页
Computer Technology and Development
基金
企业委托项目(企2007001)
关键词
形状检测
机器视觉
非接触测量
缺陷检测
拐点检测
shape test
machine vision
noncontact measurement
disfigurement detection
comer detection
