Research On The Method Of Fast Image Segmentation Based On Density

Image is considered as one of the most important information carriers in the field of computer vision, because the image is easy to obtain and contains a large mount of userful information. As a kind of low-level image processing technology, image segmentation technology is the foundation of many high-level image analysis and processing technology. The purpose of image segmentation is to extract meaningful regions from the image.The application of image segmentation technology is widely used in the fields such as product detection, intelligent transportation, medical image, remote sensing image and so on. Although a large number of image segmentation algorithms have been proposed over the past few decades, there has been no general image segmentation method which can be applied to any application scenarios, so image segmentation is still a challenging issue in image processing field.CFSFDP(Clustering by Fast Search and Find of Density Peak) algorithm is a kind of clustering algorithm based on density, which has obvious advantages compared to other clustering algorithms. However, due to the characteristics of this algorithm,if it is used for image segmentation directly, the computational cost is too high, and not suitable for large-scale real-time image analysis processing. In this paper, we will analyze CFSFDP algorithm, and some improved methods will be proposed to solve the problems of this algorithm.Superpixel is a meaningful atomic region which contains adjacent pixels that have some characteristics in common, which not only preserve the important content of the image segmentation, but also can not destroy the boundary information of the original image, so it can greatly simplify the representation of an image and reduce the complexity of subsequent image processing tasks.This paper proposes a fast image segmentation method which is combined with the advantages of these two technologies. Firstly, the SLIC(Simple Linear Iterative Clustering) algorithm is used to generate a superpixel graph. Then the CFSFDP algorithm was applied to get the final image segmentation result from the superpixel graph.A large number of experiments proved that, compared with other image segmentation methods, the proposed method has higher accuracy and efficiency.