论文部分内容阅读
在对常用图像分割质量评价方法及度量参数进行整理和研究的基础上,将均值、标准差、信息熵、空间频率、平均梯度等定量评价参数应用到土微结构图像阈值法分割质量的评价中来,对双峰阈值法、Otsu法、迭代法以及最大信息熵法分割图像的效果进行了评价。实验结果表明,针对本文图像,双峰阈值法分割效果较为理想。
On the basis of sorting out and researching the commonly used methods of image quality assessment and the measurement parameters, the quantitative evaluation parameters such as mean value, standard deviation, information entropy, spatial frequency and average gradient are applied to the quality assessment of soil microstructure image thresholding The bisexual threshold method, the Otsu method, the iterative method and the maximum entropy method were used to evaluate the segmentation results. The experimental results show that the bivariate threshold method is more effective for the image segmentation.