Multi-view three-dimensional display image synthesis method

Abstract

The invention discloses a multi-view three-dimensional display image synthesis method. The method comprises the following steps: setting the number of the images produced via an N view three-dimensional display in each view as M, and setting the number of overlapped images as Y; calculating a depth map through an image generating device based on a two-view image signal according to a formula X=N x (M-Y) + Y; estimating according to the depth map DIBR (depth image based rendering) to obtain X images; dividing M images generated in each view into M parts through an image synthesis device, acquiring 1 / M of each image from left to right, and combining 1 / M of each image from to obtain a complete image; and filling the complete image generated in each view into a pixel point corresponding to the N view three-dimensional display through the image synthesis device according to the pixel filling information of the N view three-dimensional display. According to the multi-view three-dimensional display image synthesis method, the images in M angles are inserted into single view, thus the angles of the images in each view of a three-dimensional display are increased; the number Y of the overlapped images is set, so that the images in double image state are the same image, thus the influence of the double images to the quality of the image is reduced.

Description

technical field [0001] The invention relates to the field of display image synthesis, in particular to a multi-view stereo display image synthesis method. Background technique [0002] The current multi-view stereoscopic display mainly uses the parallax of human eyes to make the left eye and right eye see different images to produce a three-dimensional effect, such as figure 1 shown, figure 1 It is a three-dimensional display screen with six viewing angles. In the existing technology, only N different stereoscopic images can be inserted on a stereoscopic display screen with N viewing angles, that is, only one stereoscopic image can be inserted in one viewing angle, so as to present a three-dimensional effect, such as figure 2 As shown, the method is to use two 2D images to synthesize a depth map, and then use one of the 2D images and the depth map to cooperate with the DIBR (Depth Image Rendering) method to generate 6 images with different perspectives, namely P1, P2, and...

AI Technical Summary

Problems solved by technology

Only a certain angle of the image can be displayed in the same viewing angle, so that the angle of the overall image presented by the stereoscopic display is also limited; and the current stereoscopic display is mainly to add gratings or lenses to the 2D display to make the human eyes see Different pixels cause parallax, but this method can only see a better image at the best viewing distance. If it is farther or closer to the viewing position, Crosstalk (ghosting) will occur, that is, it will be blocked Interference from pictures from other viewing angles, thus affecting the user's viewing effect

Images