基于马氏距离的荔枝图像分割设计方法

在基于颜色特征和聚类的马氏距离算法基础上，利用MATLAB软件设计了人机交互式可视化计算机图像分割程序。运用该程序，对荔枝照片进行了图像分割，分离出复杂背景下的荔枝果实图像。结果表明：该程序能较好地解决复杂彩图中目标图像的分割问题，可以消除因不慎而修改源代码的风险；通过数据分析器和滤波器能准确的对采样数据进行判断以及对含噪声的目标图像作进一步处理。

Computer Program Based on the Mahalanobis Distance for Image Division of Lichee Pictures

This paper presents a computer program with GUI(Graphical User Interface)that can divide the elements in a color picture.The program was made with MATLAB by using the Mahalanobis Distance algorithm based on color characteristic and color clustering.The photos of lichee fruits and leaves were used to illustrate the practicability by running the program the elements(mainly the fruits,branches and leaves) were separated and the lichee fruits were extracted forming the complicated background of these pictures.The results indicated that this program could effectively solve the problem of separating the target elements from a complicated color picture and avoid the risk of mistakenly modifying the source code.In addition,the Data Analyzer and the Noise Filter of this program made it possible to make the veracity of the sampling data and to further dispose the object image with noises.