Significance detection method based on global and local contrast

Abstract

The invention discloses a significance detection method based on global and local contrast, and the method is based on three significance operations: 1, CESC: considering the scarcity of surrounding image patch blocks; 2, CSC: extending a CESC algorithm through considering the relative position of a central image patch block with the surrounding image patch blocks; 3, GC: calculating the significance through employing a patch of a contrast image of the whole image. Finally, the three contrast images are combined. In an RGB color space, a detection result, compared with a result obtained through a conventional method, is easier and more effective. A saliency map is more attached to a result obtained by a human eye visual perception system. The method is high in quality of significance detection, and the impact on the detection result from the physical condition, optical condition and color difference of the image is small. The method is not limited by a sample, and is more suitable for actual application.

Description

technical field [0001] The invention relates to a multi-scale context-based saliency detection method, which belongs to the technical field of computer communication. Background technique [0002] In the 21st century, with the rapid development of computer technology and artificial intelligence and the continuous improvement of related theories, digital image processing technology has been widely valued in many fields. For people's better and more convenient life, the research of image processing is very important. Among them, saliency detection is an important task in computer vision and image processing. The purpose of image salient region detection research is to obtain high-quality saliency maps, which reflect the saliency of different regions in an image. Using the saliency map, the salient regions in the image can be quickly located and processed, so as to achieve the purpose of simulating the saliency of human vision by computer. [0003] However, there are still ma...

AI Technical Summary

Problems solved by technology

Due to the incomplete research on visual saliency in biology, many conclusions are speculative. Goferman et al. proposed a context integration method, which broke through the biological model and simplified the calculation method. However, the quality of saliency detection is not high.

The multi-scale method proposed by Achanta et al. adopts the local contrast calculation method, and the detection results are easily affected by factors such as complex colors in the image, object textures, and changing environmental backgrounds, making the saliency map obtained by known methods Inability to accurately highlight the foreground of the image

Therefore, the potential of image saliency information cannot be brought into full play

The histogram-based calculation method proposed by zhai and Shah uses the global contrast calculation method. When the color difference between the foreground and the background is not large, it will cause misjudgment of the salient area and cannot clearly indicate the outline of the object. This makes the saliency map The application effect is relatively poor

[0004] At the same time, the above existing methods are not ideal for saliency detection in complex situations such as image background and foreground are similar.

Existing methods only focus on a single contrast, and the detection performance is far from meeting the needs of practical applications

Images