摘要
研究了动态鸭蛋图像预处理及图像颜色特征参数的提取算法。其中预处理包括去图像边缘毛刺和蛋壳污点信息;提取鸭蛋图像颜色特征参数需要通过鸭蛋轮廓识别和圆形搜索算法实现,对于鸭蛋图像的轮廓识别是以彩色图像的红色分量R来判定图像边界点,并以此计算出鸭蛋图像蛋内像素和,用圆形搜索法求出蛋芯颜色特征区域并提取图像特征颜色参数。试验结果表明,通过该算法,得到的蛋重模型相关性达到0.95,建立的颜色模型相关性达到0.92。
In this paper, one color eigenparameter algorithm and one method of duck-egg dynamic images preprocessing which includes removing burrs from duck-egg image edge and removing the dirty information were studied. The color eigenparameter algorithm includes duck-egg images egde detection and circular searching algorithm. Using red color of RGB, the image＇s edge and sum pixels inside duckegg images were obtained. Through circular searching algorithm, the color eigenparameter was obtained. The experiments show that using this method,the weight math module＇s relative coefficient is 0. 95, the color math module＇s relative coefficient gets to 0. 92.
出处
《华中农业大学学报》
CAS
CSCD
北大核心
2005年第5期512-515,共4页
Journal of Huazhong Agricultural University
基金
国家星火计划项目(2001EA760026)
湖北省重点科技攻关项目(20002P0603)资助
关键词
鸭蛋
机器视觉
图像处理
圆形搜索算法
duck-egg
machine vision
image processing
circular searching algorithm
