Image segmentation method based on visual salient region and active contour

Abstract

The invention discloses an image segmentation method based on a visual salient region and an active contour, and belongs to the field of mode recognition and intelligent information processing. The method comprises the following steps: carrying out preprocessing operation on an original image by adopting an improved visual salient region detection method, automatically setting an initial contour, designing a self-adaptive sign function by using acquired target prior information, weighting an optimized LoG energy item, and constructing the optimized LoG energy item weighted by the self-adaptive sign function; fully considering local space and gray level change information, and improving a local energy item; and finally, fusing the global energy item and the improved local energy item in a linear mode, and proposing a new hybrid active contour segmentation model based on the region. According to the method, the target position can be quickly positioned in the image containing the complex background, and the consistency and integrity of the target contour are kept.

Description

technical field [0001] The invention relates to an image segmentation method based on visually significant regions and active contours, and belongs to the field of pattern recognition and intelligent information processing. Background technique [0002] Image segmentation is the basic research content of real-time image processing. As a preprocessing operation of target extraction, recognition and tracking, the quality of image segmentation results directly affects the processing results of subsequent steps. [0003] Usually, in the process of image acquisition, it is easily affected by the equipment and the external environment, and the acquired images often have the problem of uneven distribution of gray levels and blurred boundaries. Therefore, image segmentation faces huge challenges. The image segmentation method based on the active contour model utilizes the principle of energy minimization, which can provide smooth contours for object segmentation. Since such models ...

AI Technical Summary

Problems solved by technology

[0007] In order to solve the problem that the traditional active contour algorithm cannot automatically set the initialization contour, the curve evolution is easy to fall into local extremum, and it is difficult to segment images with uneven gray scale and weak edges, the present invention provides an image based on visually significant areas and active contours Segmentation method, which firstly improves the existing visually salient area detection algorithm, designs a compact contrast measure combined with cluster-based compactness, obtains more accurate target prior information, automatically sets the initialization contour curve and optimizes the symbol function of the LoG energy term ;Use the adaptive sign function to weight the optimized LoG energy item, and propose an optimized LoG energy item weighted by the adaptive sign function based on the global information of the image, which can effectively avoid the model from falling into local extremum during the evolution process and improve the accuracy of model segmentation; Introduce the combination of local gray level change information and spatial information, improve the local energy item, improve the sensitivity of the model at weak edges, and make the obtained target contour more complete; finally, the adaptive sign function weighted LoG Energy Term and Improved Local Energy Term Applied to Image Segmentation

Images