Image segmentation method based on cluster ensemble

Abstract

The invention discloses an image segmentation method based on a cluster ensemble in the field of computer vision and pattern recognition. According to the image segmentation method, different coarse segmentation results and superpixels used for follow-up processing can be generated through three segmentation modes; then, distance attributes between different superpixels and annotation attributes of an image in the coarse segmentation results are integrated through a superpixel clustering integration method; in this way, the segmentation result of the image is improved, complex calculation is avoided, and color and texture attributes of the pixels are manifested.

Description

technical field [0001] The invention belongs to the field of computer vision and pattern recognition, and in particular relates to a new non-supervised image segmentation method based on image multi-information clustering fusion. Background technique [0002] In the field of computer vision, image segmentation is an important step in image preprocessing. Since the current computer itself does not have the high-level comprehension ability of human beings and other organisms, and only processes discrete data, it is very necessary to use preprocessing such as image segmentation to process images. The result of image segmentation plays an important role in the subsequent processing. The subsequent image processing involved in image segmentation includes, for example, target detection and recognition, image classification, target tracking, and image compression and reconstruction. [0003] At present, there are many kinds of image segmentation methods, using various methods, suc...

AI Technical Summary

Problems solved by technology

Since only a single attribute of the image is used to cluster each pixel in the image, these segmentation methods are often suitable for image segmentation for a specific purpose, and do not have good applicability

Or a more complex modeling method is used, which makes the efficiency of image segmentation lower and the segmentation results are also limited.

Images