摘要
基于机器视觉技术设计了鸡胴体表面污染物在线检测及处理系统。通过'工业相机+滤光片+计算机'的方式在线采集鸡胴体表面波长500 nm和710 nm的特征图像,采用中值滤波、灰度增强进行图像预处理,然后采用Otsu法自动确定阈值并获得二值化图像,再对图像进行腐蚀、膨胀、空洞填充以及异或操作,分割得到鸡胴体表面污染物区域并以此判断鸡胴体表面是否存在污染物,然后对鸡胴体表面的污染物进行喷淋处理。试验结果显示,利用该系统进行鸡胴体表面3种污染物(盲肠粪便、血液、胆汁)检测,检测总体平均正确率为90.5%,表明该系统可实现鸡胴体表面污染物的在线检测和正确识别。
A real-time detection and processing system for contaminants on chicken carcass surface was designed based on machine vision technology. The surface feature images in the wavelengths 500 nm and 710 nm of chicken carcass were acquired by means of industrial camera, filter and computer respectively. These images were firstly processed with the median filtering and gray enhancement. Then the binarized images were obtained by using Otsu automatic thresholding. Then, the corrosion, expansion and void filling operations were conducted on these images, and the contamination areas on chicken carcass surface were segmented. On the basis of this, if contaminations （cecal fecal, blood, bile） on the surface of chicken carcass were distinguished, the spray treatments for these contaminations were carried out. The experimental results showed that the total average correct ratio was 90. 5 % when this system was used to detect three contaminations on the surface of chicken carcass, implying that this system could be used for online detection of contaminants on chicken carcass surface with high detection accuracy and reliability.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2015年第9期228-232,共5页
Transactions of the Chinese Society for Agricultural Machinery
基金
江苏省科技支撑计划(社会发展)资助项目(BE2014708)
关键词
鸡胴体
表面污染物
机器视觉
在线检测
图像处理
Chicken carcass Contaminants on the surface Machine vision Online detection Imageprocessing
