改进C-V模型在YCbCr空间下的羊体图像分割
Sheep image segmentation by an improved Chan-Vese model in YCbCr space

针对羊体图像背景复杂、分割难以及不同光照条件干扰羊体图像的问题，采用一种基于YCbCr空间改进C-V主动轮廓模型的分割方法，对具有复杂背景的羊体图像分割进行研究。结果表明：1）根据羊体图像的颜色特点，对羊体图像进行从RGB空间到YCbCr空间的转换能克服拍摄环境中光照对羊体的影响；2）利用手动勾画羊体的粗略轮廓构造预处理水平集，对其内部、外部以及边界进行划分后可以演化羊体图像的轮廓。试验证明改进C-V模型能对复杂背景下的羊体图像进行准确分割，分割结果能够应用到后续羊体测量点的识别中。
To solve the problem that the object was difficult to segment from complex background in the sheep images and the illumination interference on images, an improved C-V active contour model based on YCbCr space was used to study sheep images segmentation in complex backgrounds. The results showed that:1) According to color feature of the sheep images, conversion from RGB space to YCbCr space was done for each pixel, which was necessary to overcome illumination factors influencing sheep images. 2) A new level set was constructed by sketching the rough outline of sheep manually in order to distinguish an outer region from inner region, and the curve evolution was halted at the object contour border. It was proved that the improved C-V model could segment the sheep images in complex background and the segmentation results could be applied to succeeding sheep measurement point identification.