Abstract: For the purpose of measuring the micro changes of morphological parameters of melon, a new LCV（Local Chan-Vese） model algorithm based on prior shape focusing on segmentation of melon image is proposed. Local boundary information image of melon is acquired through mathematical morphology algorithm, and LCV model is constructed according to the form of traditional CV model. Prior shape which is obtained by mathematical morphology and spline interpolation algorithms is constructed, and the prior shape is integrated into LCV model functional through a shape comparing function. LCV model based on prior shape is constructed. Compared with traditional edge detection and segmentation algorithms, more ideal boundary information can be obtained by this new algorithm.

Key words: melon morphology, micro-varying measurement, image segmentation, prior shape, Local Chan-Vese（LCV） model

摘要： 以甜瓜形态的微变测量为目标，提出了基于先验形状的LCV（Local Chan-Vese）模型算法。利用形态学方法获得甜瓜的局部信息图像，仿照传统CV模型的拟合形式，建立了LCV能量模型；利用形态学方法及spine插值算法获取甜瓜的先验形状，将其通过形状比较函数嵌入到LCV模型的能量函数中，形成了新的基于先验形状的LCV模型。相比传统边缘检测和图像分割算法，该算法更易于提取出理想的边缘信息。

关键词: 甜瓜形态, 微变测量, 图像分割, 先验形状, 局部Chan-Vese分割（LCV）模型