摘要
在图像分割提高精度问题的研究中,要从图像中提取感兴趣的区域。由于图像模糊或者蜕化图像背景信息融合在一起,导致难以区分,传统的阈值图像分割算法容易造成分割效果不清晰。为解决上述问题,提出了一种新的快速有效的两级阈值结构图像分割算法,采用用迭代算法对图像进行单一阈值分割,在每次迭代过程中以图像均值为依据,对图像进行均衡化处理;在基于全局分割的基础上,在局部范围内根据噪声的统计特性对图像进行去噪处理。仿真结果表明,提出的两级阈值分割算法能快速有效地分割图像,不仅可以得到了比较高的分割精度,还大大减少了计算量,一定程度上能够改善图片分割的效率和质量。
The binary segmentation of color images improves the accuracy of the problem. Degenerate image makes it difficult to distinguish image from the background, and the traditional image segmentation algorithm for anti - noise ability, eventually leading to the defect of low segmentation accuracy. A fast and efficient image segmentation was propose based on two - level threshold structure. Firstly, an iterative algorithm was used in single image segmentation, , based on the mean of image in each iteration, the equilibrium processing was carried out. Based on global segmentation, the image denoising was carried out in the local area according to the noise statistical properties. Simulation results show that the proposed two - level threshold segmentation algorithm is fast and effective. It can not only get higher accuracy of segmentation, but also greatly reduce the computation, and improve the efficiency and quality of image segmentation to some extent.
出处
《计算机仿真》
CSCD
北大核心
2012年第2期253-256,共4页
Computer Simulation
关键词
两级阈值
二值化
图像处理
图像分割
Level threshold
Binary
Image processing
Image segmentation
