利用图像处理技术自动估算猪胴体参数*

采用图像阈值分割和彩色分割算法，以及轮廓跟踪和拐点提取方法，成功实现猪二分体图像特征提取,用VC++语言实现自动估算猪胴体图像参数。研究结果表明迭代阈值分割和彩色分割算法能正确地分割二分体图像，距离求索法能够对拐点正确提取，进而确定特征点。用获取的图像特征与胴体实测值建立的关系模型有较好的线性关系（p<001），表明图像处理技术可以快速无损地估算猪胴体参数，评价猪肉质量。

Evaluating Parameters of Hogs's Carcass Using Image Processing Technology

The threshold segmentation and color segmentation algorithms are adopted,the image features of half carcasses is obtained using the track outline and inflexion method,(VC ) is then used to calculate the fat thickness,area of buttocks automatically.The results indicate that the repeatedly threshold segmentation and color segmentation algorithms could segment the half carcass image; The distance track method could correctly seek inflexion,then confirm the image features.Fat thickness,area of buttocks of hog's half carcass could be estimated automatically using the established models in this study with image processing technology,minimizing human errors and contamination,and improving efficiency in assessing pork quality.