A stereoscopic video saliency detection method based on binocular multi-dimensional perception characteristics

Abstract

The invention relates to a stereoscopic video saliency detection method based on a binocular multidimensional perception characteristic. A traditional model method cannot effectively detect a salient region of a stereoscopic video. The method comprises the steps of salient feature extraction and salient feature fusion; the step of salient feature extraction is as follows: respectively performing saliency calculation from view information of the stereoscopic video of three different dimensions of space, depth and motion, including two dimensional static salient region detection, depth salient region detection, and motion salient region detection; and the step of salient feature fusion is as follows: using a global nonlinear normalized fusion strategy to perform fusion on an acquired salient feature map of three different dimension, so as to acquire a stereoscopic video salient region. According to the stereoscopic video saliency detection method based on the binocular multidimensional perception characteristic provided by the invention, the computational complexity is low, the quality of an acquired stereoscopic video saliency map is high, and the method can be directly applied to 3D video compression, 3D quality assessment and object identification and tracking and other engineering fields.

Description

technical field [0001] The invention belongs to the technical field of video image processing, and in particular relates to a stereoscopic video saliency detection method based on binocular multi-dimensional perception characteristics. Background technique [0002] Three-dimensional (Three-Dimension, 3D) video can bring viewers an immersive experience and higher fidelity due to the parallax between the left and right viewpoint images. It is a new generation of video service technology that is currently being developed. However, research on human vision has shown that due to the focusing function of the eyeball, the human eye cannot simultaneously perceive near objects and distant objects in a 3D video, and must focus on a certain area, resulting in a stronger selectivity of human 3D vision than 2D vision. , showing that its regional saliency is more prominent on 3D video. The 3D video saliency calculation model has important guiding significance for the calculation and reco...

AI Technical Summary

Problems solved by technology

[0003] Most of the traditional saliency detection models use different algorithms for saliency detection based on the spatial characteristics of the image such as color, brightness, direction, and texture. However, these traditional model methods cannot effectively detect the salient regions of stereoscopic videos. On the one hand, Since most of the traditional detection models do not calculate the significant characteristics in the time domain, the motion between adjacent frames is one of the important features that affect the visual attention of the human eye, and the commonly used method for detecting motion features is the frame difference method. , background modeling method and optical flow method, etc.

The frame difference method is relatively simple, but the accuracy rate is low. The background modeling method is greatly affected by the background model, which will lead to unstable detection, while the algorithm complexity of the optical flow method is high. On the other hand, the traditional detection model does not calculate the depth information. For the impact of the salient characteristics of stereoscopic video, the detection is not accurate enough, because the depth information reflects the distance of the object from the human eye, which is one of the important perceptual characteristics of stereoscopic video

Images