Image significance detection method combining color and depth information

Abstract

The invention discloses an image significance detection method combining color and depth information. The method comprises the following steps: performing superpixel segmentation on a to-be-detected color image, calculating a region contrast image in each segmented area through combining depth and color features, and obtaining a depth prior image and a direction prior image by utilizing depth information; integrating the region contrast image, the depth prior image and the direction prior image, and obtaining a contrast image integrated with prior information through calculation; performing overall optimization on the contrast image integrated with prior information: executing the normal inner product weighted webpage ranking algorithm, selecting an area with high confidence coefficient as a sampling area, designing an image restoration problem based on a Markov random field model, and solving to obtain a final significance detection image. According to the invention, the influence of the depth and direction information on significance is explored, and compared with the existing image significance detection method combining color and depth information, the method provided by the invention achieves a better effect.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to an image saliency detection method combining color and depth information. Background technique [0002] Image saliency detection is one of the hot topics in the field of computer vision and pattern recognition. The study found that the human visual mechanism can always quickly extract the important and interesting regions in the image, and then analyze and process these regions, but basically does not process the remaining insignificant regions in the image. This principle provides a lot of inspiration for researchers in the field of computer vision, that is, it can detect the salient areas in the image, extract the salient objects in the image for subsequent processing, save the time for processing the whole image, and greatly improve the efficiency of image processing . Therefore, image saliency detection can be widely used in image segmentation, object r...

AI Technical Summary

Problems solved by technology

Houwen Peng et al. use color and depth features to calculate local contrast, then use the region growing algorithm to cluster salient areas, and finally use the prior knowledge of object positions to improve the results. This method only uses depth information and depth calculations. Normal information is used as an image feature component other than color, but the detection effect of objects with inconspicuous color information is not good

Ran Ju et al. used depth information to calculate the anisotropic center-surrounding difference of the disparity map to achieve saliency detection, but did not combine it with color information

Yuzhen Niu et al. calculated the global parallax contrast, and achieved saliency analysis based on the assumption that salient objects are usually located in the visual comfort zone of stereo images, but the impact of parallax or depth on the saliency of objects is still not deep enough.

Images