Optimized and accelerated depth map super-resolution interpolation method

Abstract

The invention discloses an optimized and accelerated depth map super-resolution interpolation method, which comprises the following steps of: pre-defining a pre-sampling kernel, storing a coordinate relationship between each pixel point in a window and four known pixel points of the window in the pre-sampling kernel, and optimizing floating point type numerical calculation into integer numerical calculation, the purpose of optimizing the speed is achieved on the premise that the precision is not lost, and the efficiency of the interpolation algorithm is improved. Different interpolation strategies are adopted at the boundary, the pyramid noise problem at the boundary is eliminated in the reconstruction process, denoising processing does not need to be carried out on the reconstructed depth map again, and the calculation amount and time consumption are reduced. In order to eliminate the problem that a large amount of noise exists in a cavity area of a depth map after super-resolution reconstruction, denoising, cavity detection and cavity filling are carried out on a low-resolution depth map, namely an original depth map, before reconstruction, and the noise and cavity problems of the low-resolution depth map are eliminated.

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 image, which is used to record the geometric information of the scene. In recent years, the RGB-D data format combined with texture maps and depth maps 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 by mobile devices. However, the acquired quality of depth maps, which are an important part of modern multimedia, is not satisfactory. At present, the mainstream time-of-flight-based TOF and depth sensor technology based on structured light technology obtain depth maps with low resolution and are susceptible to interference from ambient light noise, so it is necessar...

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