一种基于复杂背景下的昆虫彩色图像分割方法

A NATURAL INSECTS COLOR IMAGE SEGMENTATION ALGORITHM

根据复杂背景下昆虫彩色图像的特点,采用了一种融合颜色和空间信息的JSEG分割方法。首先对图像进行颜色量化,然后使用区域增长法合并较小区域从而实现图像分割。在实验中,利用JSEG分割方法分别对原图像和加白噪声后的图像进行分割,结果表明,JSEG算法对昆虫图像中的噪声不敏感,可以高效地实现复杂背景下昆虫彩色图像的分割。

A JSEG algorithm based on color and space information is presented for natural insects color image segmentation. Firstly, color quantization is performed on an image. Then, a region connected procedure is used to connect similar neighbors into bigger ones. In the experiments, both original images and noised images are segmented by JSEG method. Experimental results show that JSEG algorithm is insensitive to noise, and it is able to segment natural insects color images effectively.