结合C-V模型水平集与形态学的彩色树木图像分割

树木图像分割是一种从图像中把树木与周围背景完整分离的技术，是计算机仿真学科在林业应用方面的核心内容，也是计算机视觉方向的研究热门，为林业应用提供一定的技术支持。根据树木图像含有分裂、合并、形成尖角等相对比较复杂形状的特点，首先对图像运用基于C-V模型水平集的计算，通过计算活动轮廓长度和差异量来判断迭代收敛的情况，待迭代稳定后对其进行形态学后处理操作，将某些过分割区的细密纹理和噪声剔除，从而得到全局最佳优化的图像分割效果。为彩色树木图像的分割提供一种更为有效的方法。

Color Tree Image Segmentation Method Integrating A Set of C-V Plane Models with Morphological Processing Operation

Tree image segmentation is a technology that separates trees from their surrounding landscapes. It is the core technology in virtual reality and computer simulation of forestry applications, and is also one of the focus areas in machine vision to provide basic data and technical support for the application of computer technology in forestry. Because of the characteristics of complex shapes of the tree images, which contain fission, mergence and sharp corner formation, and so on. This paper used a set of C-V plane models, by calculating the active contour length and the difference quantity to judge the iterative convergence, waiting until the iterative becomes stable, and then combined with later morphological processing operation, the fine texture and noise in the non-target sections of the primarily segmented image could be eliminated, so that the overall optimized image segmentation effect could be obtained in a rapid and accurate way. This result might provide color tree image segmentation with an effective method.