Research On Algorithms For Image Segmentation Based On Fuzzy Clustering With The Spatial Information

Image segmentation is one of the most essential parts of forming digital image processing,which is also one of the core technologies of smoothly carrying out image analysis, understanding image and depicting image. Image segmentation technology refers to a technology and process that separates out the target area required by people from an image, and it is widely used in many fields.Fuzzy cluster analysis occupies a crucial position in image segmentation algorithm. Fuzzy clustering has been widely used in image segmentation, and it achieves good segmentation results because the image itself has a lot of imprecision and uncertainty, while fuzzy theory has a very powerful depicting ability with regard to this, among which fuzzy c-means clustering is a clustering algorithm that is particularly well-known by people.A key study is carried out in terms of the FCM algorithm, and the main content is as follows:(1)Research on the FCM image segmentation algorithm and its deficiencies.FCM algorithm has the advantage of good convergence, and its depicting ability for the uncertainty of things is excellent too. However, FCM algorithm still has some shortcomings: it takes long time for large data models arithmetic, and the edge of the image gained by segmentation is not smooth; it is hard to achieve the segmentation of the gray level similar region between the target and non-target region.(2)Research on the FCM image segmentation algorithm based on quadtree and curve fitting.The improved quadtree is applied in image segmentation for fuzzy C-means clustering algorithm to solve the problem of long time consuming.The algorithm firstly obtains the image of quadtree, using the quadtree node collection as a sample set of fuzzy clustering algorithm and reducing the sample space, then uses FCM algorithm to realize the image segmentation; With the jagged edges for the image segmentation brings, it uses curve fitting algorithm for image edge smoothing.The algorithm is applied in standard photographer image and liver magnetic resonance image. Experiment results show that this algorithm improves the segmentation speed of FCM and it has higher accuracy of segmentation of medical image and provides a basis for further diagnosis and data analysis.(3)Research on the FCM image segmentation algorithm based on cellular automata.The cellular automata is applied in image segmentation for fuzzy C-means clustering algorithm. In order to improve the segmentation effect of FCM algorithm on the gray level area between the object and background in the image, this experiment firstly enhances the image’s contrast by usingcellular automata, and the state transition function uses the Moore neighborhood and the same evolution rule, and then to achieve the edge segmentation of the image based on FCM algorithm.The algorithm is applied to general photos and medical images, and the experimental result shows that the algorithm improves the accuracy of the image segmentation, which provides the foundation for the subsequent analysis and research. The experiments verify the effectiveness of the improved algorithm.