论文部分内容阅读
Based on the maximum fuzzy entropy principle,the edge image with low contrast is optimally classified into two classes adaptively,under the condition of probability partition and fuzzy partition.The optimal threshold is used as the classified threshold value,and a local parametric gray-level transformation is applied to the obtained classes.By means of two parameters representing,the homogeneity of the regions in edge image is improved.The excellent performance of the proposed technique is exercisable through simulation results on a set of test images.It is shown how the extracted and enhanced edges provide an efficient edge-representation of images.It is shown that the proposed technique possesses excellent performance in homogeneity through simulations on a set of test images,and the extracted and enhanced edges provide an efficient edge-representation of images.
Based on the maximum fuzzy entropy principle, the edge image with low contrast is optimally classified into two classes adaptively, under the condition of probability partition and fuzzy partition. The optimal threshold is used as the classified threshold value, and a local parametric gray-level transformation is applied to the obtained classes.By means of two parameters representing, the homogeneity of the regions in edge image is improved.The excellent performance of the proposed technique is exercisable through simulation results on a set of test images. It is illustrated how the extracted and enhanced edges provide an efficient edge-representation of images. It is shown that the proposed technique possesses excellent performance in homogeneity through simulations on a set of test images, and the extracted and enhanced edges provide an efficient edge-representation of images.