Image Segmentation Algorithm Based on Non-Local Information and Subspace for Fuzzy C-Ordered Mean Clustering

Aiming at the problem that the fuzzy C-ordered means algorithm does not consider the spatial information of the image, which makes it difficult to effectively segment the noisy image, a fuzzy C-ordered means clustering algorithm based on non-local information and subspace is proposed. Firstly, the non-local spatial information of the current pixel is extracted by using the pixels of the given similar neighborhood structure in the image. Secondly, the typicality of each pixel is calculated and sorted, and in each iteration update the typicality of pixels and to solve the problem of misclassification caused by similar classes during the clustering process. Finally, the concept of subspace clustering is introduced to allocate appropriate weights to different dimensions of the image, thus improving the segmentation performance of color images. The experimental results on the noisy synthetic images and public datasets BSDS500, MSRA100 and AID show that the partition coefficient, partition entropy, segmentation accuracy and normalized mutual information average of the proposed algorithm achieves 95.00%, 6.66%, 98.77% and 95.54%, respectively, which are better than similar methods in comparison.