A Depth Map Generation Method Based on User Perception Optimization

Abstract

The invention discloses a method for generating a depth map based on user perception optimization, which includes the following steps: input a plane or 2D image, use a neural network algorithm to generate a rough depth map, use the neural network algorithm or manually mark out the region of an object in the image And generate a segmentation map, the region of the object is a segmentation block; calculate the feature of the image, and calculate the attention score in conjunction with the feature of the image; carry out the energy function design through the rough depth map and the attention score, and obtain the optimized depth map; The optimized depth map is iteratively adjusted to resolve the contradiction of the boundary depth difference of the segmented blocks, and the resulting depth map is output. The present invention does not need a stereo camera, can input a common plane picture, generate a depth map with better effect, realize three-dimensional image display, and reduce the cost of three-dimensional image output; adopt the method based on user perception optimization, and solve the problem of The depth map produced by the stereo map generation method is blurred and the effect is not good.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method for generating a depth map based on user perception optimization. Background technique [0002] Naked-eye stereoscopic images provide an additional depth dimension, which is expected to better display data, structures, etc., and have been used in medical diagnosis, visualization, and other fields. For example, naked-eye stereoscopic images can help doctors better understand the structure of the human body and speed up operations during surgery. Studies have shown that compared to flat displays, stereoscopic images can provide a more immersive environment and stimulate positive emotions. [0003] Existing methods do not take into account the cognitive characteristics of the user when viewing the naked-eye stereoscopic content, such as the user's depth resolution, the user's attention area on the image content, etc., so that the content displayed in the naked-eye...

AI Technical Summary

Problems solved by technology

This invention uses wavelet decomposition to optimize the depth map, but does not consider the user's attention and depth perception characteristics

Images