An Optimized and Accelerated Depth Map Super-resolution Interpolation Method

Abstract

The invention discloses an optimized and accelerated depth map super-resolution interpolation method. By pre-defining the pre-sampling kernel, the coordinate relationship between each pixel point in the window and four known pixel points of the window is stored in the pre-sampling kernel. And optimize the floating-point numerical calculation into integer numerical calculation, achieve the purpose of optimizing the speed without losing precision, and improve the efficiency of the interpolation algorithm. The present invention adopts different interpolation strategies at the boundary, eliminates the pyramid noise problem at the boundary during the reconstruction process, does not need to denoise the reconstructed depth map again, and reduces the amount of calculation and time consumption. In order to eliminate the problem of a large amount of noise in the hole area of ​​the depth map after super-resolution reconstruction, the present invention chooses to perform denoising, hole detection and hole filling on the low-resolution depth map, that is, the original depth map, before reconstruction, so as to eliminate the low-resolution depth map. The problem of noise and holes in the rate depth map.

Description

technical field [0001] The invention relates to the technical field of image data processing, in particular to an optimized and accelerated depth map super-resolution interpolation method. Background technique [0002] Compared with traditional images, a depth map is a special kind of image, which is used to record the geometric information of the scene. In recent years, the RGB-D data format combining texture map and depth map has been widely used in high-tech fields such as virtual reality and 3D reconstruction. With the continuous improvement of imaging technology, high-quality texture maps have been achieved on mobile devices. However, the acquisition quality of depth maps, which are an important part of modern multimedia, is not satisfactory. At present, the depth map obtained by the mainstream TOF based on time of flight and depth sensor technology based on structured light technology has low resolution and is easily interfered by ambient light noise, so it is necess...

AI Technical Summary

Problems solved by technology

Depth map super-resolution reconstruction is essentially a process of upsampling interpolation from low-resolution images to high-resolution images. However, the interpolation algorithms in the prior art are floating-point data, which requires a large amount of calculation and slows down the overall reconstruction efficiency.

Images