摘要
通过计算机视觉系统获取鸭蛋表面的图像,利用颜色特征参数和灰度阈值方法对图像进行分割。提取分割后的裂纹区域和噪声区域的6个几何特征参数,对图像进行识别,裂纹识别率为93.392%,噪声识别率为93.602%。
Images of the duck egg shell surface which obtained by computer vision system were portioned and analyzed according their color feature parameters and grayscale threshold values. Based on 6 geometric parameters of the noise and crack regions, the crack and noise recognition rates were 93.392% and 93.602%, respectively.
出处
《湖北农业科学》
北大核心
2009年第12期3152-3154,共3页
Hubei Agricultural Sciences
基金
国家"863"计划项目(2007AA10Z214)
湖北省科技攻关计划项目(2007AA201C64)
关键词
鸭蛋
机器视觉
裂纹
检测
duck eggshell
machine vision
crack
detection
