基于S-粗集的图像阈值分割方法

介绍了S-粗集的概念, 结合其动态迁移特性给出了可以适应复杂背景和含噪环境的图像S-粗集表示模型, 使静态目标可以将"不好"特性像素点迁移出去. 利用粗糙熵平衡目标和背景粗糙度对边界的影响, 提出一种更具适应性的 图像阈值分割算法. 为了适应离散点的迁移, 同时避免粒度大小的选择, 结合包含度概念给出了图像变精度S-粗集表示模型, 利用精度参数来 控制调节获取最佳分割阈值, 实现目标提取. 仿真实验表明, 所提出算法具有更好的图像分割效果.

Thresholding image segmentation based on S-rough set

The concept of S-rough set is introduced to represent images as S-rough sets for fitting complex background and noisy environment by combining migration characteristics. This model can remove pixels with singular nature from original object dynamically. A more adaptable technique of image thresholding segmentation is proposed by establishing a rough entropy to compromise the values of object roughness and background roughness. In order to adjust the transfer of singular points and avoid choice of granule size, images are further described by variable precision S-rough sets in combination with the notion of inclusion degree to extract object from background. Experimental results show that the proposals have better performance of thresholding segmentation of noised images.