摘要
牛肉眼肌区域大理石花纹的丰富程度是牛肉分级的重要指标之一。采用工业相机采集多幅牛肉眼肌切面图像,利用VC++图像处理技术,对图像进行平滑去噪操作,采用自适应阈值法将眼肌区域从背景中分离,然后运用数学形态学的方法以及基于区域分割的方法确定有效眼肌区域,最终通过数理统计的算法识别大理石花纹。结果表明,该技术能有效识别眼肌区域中的大理石花纹,其耗时短、识别结果准确,利于牛肉等级的准确判定。
Marbling level of beef fib-eye region is one of the important indicators in beef classification. A total of 50 fib-eye images were taken by using industrial cameras, then VC++ image processing technology was used to denoise image through smoothing operation; then the fib-eye region was isolated from background with adaptive threshold before the effective fib-eye region was determined by using mathematical and segmentation-based method. Eventually we could identify marbling through mathematical statistics method. Experimental results show that machine vision can effectively identify fib-eye region of the marbling, and its short time-consuming and accurate identification will help determine the exact quality level of beef.
出处
《食品科学》
EI
CAS
CSCD
北大核心
2011年第3期10-13,共4页
Food Science
基金
国家现代农业(肉牛)产业技术体系项目(080600231\080600232)
关键词
牛肉眼肌
大理石花纹
腐蚀
膨胀
图像分割
识别
beef rib-eye
marbling
erosion
dilation
image segmentation
recognition