基于改进的距离保持水平集的牛肉图像背长肌自动分割  

Automatic segmentation of longissimus dorsi in beef image based on improved distance preserving level set

作　　者：张彧龙 刘雅思[1] 郑丽敏[1,2] 杨璐[1] ZHANG Yulong;LIU Yasi;ZHENG Limin;YANG Lu(College of lnformation alut Electrical Engineering,China Agricultural University,Beijing 100083,China;Beifing Laboratory of Food Quality and Safety,China Agricultural University,Beijing 100083,China)

机构地区：[1]中国农业大学信息与电气工程学院,北京100083 [2]中国农业大学食品质量与安全北京实验室,北京100083

出　　处：《计算机应用》2018年第A01期213-217,共5页journal of Computer Applications

基　　金："十三五"国家科技支撑计划项目(2016YFD0401504)

摘　　要：为了改善牛肉图像中背长肌的分割效果,提出一种改进的距离保持水平集方法实现牛肉图像中背长肌区域的自动分割:根据牛肉背长肌图像的特点,结合大津法和旋转法提取背长肌切面区域,设定候选区域并找到初始轮廓曲线;根据水平集演化过程中自定义系数函数值的变化规律,使用自定义的算法控制零水平集的自动演化和停止;之后结合标记分水岭算法对图像进行加工,得到更精确的目标背长肌区域。与其他分割算法比较,该改进的分割方法效果更好。In order to improve the segmentation effect of longissimus dorsi in beef images, an improved distance preserving level set method was proposed to realize the automatic segmentation of the longissimus dorsi area in beef images. As the longissimus dorsi area is often in the middle of the longissimus dorsi image region, according to the characteristics of this position of the longissimus dorsi area, the longissimus dorsi image region was extracted using the Otsu method and a rotation method, thereby the candidate area was set and the initial profile curve was determined; According to the change rule of the custom defined coefficient function in level set evolution process, the automatic evolution and stop of the zero level set were realized by the custom defined algorithm. Combined with the marked watershed segmentation algorithm, the image was processed to obtain more accurate segmentation results. Compared with other segmentation algorithms, it is proved that the segmentation method proposed in this paper achieves the best segmentation result.

关 键 词：距离保持水平集 图像分割 牛肉图像 背长肌 

分 类 号：S823.9[农业科学—畜牧学] TP391.4[农业科学—畜牧兽医]