Image saliency target detection method

Abstract

The invention relates to an image saliency target detection method. Region segmentation of image analysis is involved. The method is an image saliency detection method based on multi-graph model priorand short connection network optimization. The method comprises the following steps of: utilizing color and position information for each input image; calculating a KNN graph model and a K regular graph model; obtaining a saliency map S1 under the KNN graph model and a saliency map S2 under the K regular graph model; carrying out pixel-level fusion on the KNN graph model and the K regular graph model; S3, obtaining an initial saliency map of the original image; and S3, optimizing the initial saliency map by using a short connection network to obtain a final saliency map Sfinal of the originalimage, and completing image saliency target detection, thereby overcoming the defects of incomplete saliency target detection and inaccurate algorithm detection when foreground and background colorsare similar in the prior art of image saliency target detection.

Description

technical field [0001] The technical solution of the present invention relates to the area segmentation of image analysis, in particular to an image salient object detection method. Background technique [0002] Image salient object detection refers to the use of computers to simulate the visual attention mechanism of the human eye to extract areas of human interest from images, which is one of the key technologies in the field of computer vision. [0003] In the prior art of image salient object detection, according to the different types of extracted image features, image salient object detection methods are divided into manual models and deep learning models. The manual model of image saliency object detection refers to calculating the salient value of the region based on image manual features, such as color, texture, position and gradient, but the low-level image manual features cannot describe the semantic information of the object, and cannot be accurate in complex sce...

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is to provide an image saliency target detection method, which is an image saliency detection method based on multi-image model prior and short connection network optimization. The method is to use color and position for each input image. information, calculate the KNN graph model and the K regular graph model, and obtain the saliency graph S under the KNN graph model 1 and the saliency graph S under the K-regular graph model 2 , and then fuse the KNN graph model and the K regular graph model at the pixel level to obtain the initial saliency map S of the original image 3 , using a short-connection network to optimize the initial saliency map S 3 , to obtain the final saliency map S of the original image final , complete image salient target detection, overcome the shortcomings of incomplete salient target detection and inaccurate algorithm detection when the foreground and background colors are similar in the prior art of image salient target detection

Images