基于优度法的图像分割性能评价

Image Segmentation Evaluation Based on Priority Degree

利用图像的模糊度衡量图像分割的质量,并设计了一种实用的非线性映射函数,将图像从空间域变换到模糊性质域。构造评价模型,结合一系列实例对算法性能进行评价。实验表明,该方法得到的评价结果反映的分割差异与人的视觉效果比较接近。

In this paper, fuzzy degree is used to measure segmentation quality. Firstly a new non-linear mapping function is proposed to change images from spatial field to fuzzy character field. Then, an evaluation model is constructed. At last, evaluation effects are obtained by using four different segmentation methods to segment nine real images. The experimental results show that the effects obtained from the new evaluation method are closed to people＇s vision effects.