摘要
首先利用牛顿二分法快速识别鸡蛋图像边缘信息;然后通过边缘信息简化图像区域面积像素和公式及图像形心坐标公式,提出自动确定长轴的方法;最后建立了鸡蛋蛋形指数和质量的预测模型。验证结果表明,边缘检测效率提高了大约20倍;实际蛋形指数与预测指数的相关程度达到0.9523;通过图像面积像素和与图像旋转体积像素和所预测的质量与实际质量的相关性达到0.9806。
The paper provided a fast method to distinguish the edge information from the egg image based on Newton Dichotomy. With the information, the author gave the new brief formulas to calculate the image element sum of the target area and the coordinate of egg figure center, as well as proposed a method for auto-search for the longest axis. A model for forecasting the relation between egg weight and its shape index was established. The results showed that the edge identification efficient was increased about 20 times via the new method, the real shape index and forecasting shape index of egg reached a correlation degree of 0. 952 3, and the real weight and the estimate weight of egg reached a correlation degree of 0. 980 6.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2007年第11期80-83,共4页
Transactions of the Chinese Society for Agricultural Machinery
基金
国家星火计划资助项目(项目编号:2001EA76002)
湖北省重点新产品计划资助项目(项目编号:2005BDS043)
华中农业大学科技创新基金资助项目(项目编号:52204-05079)
关键词
鸡蛋
机器视觉
检测
质量
预测模型
Egg, Machine vision, Detection, Weight, Prediction model
作者简介
博士生
副教授
副教授 博士生
教授 博士生导师 通讯作者
