Depth map super-resolution reconstruction method based on convolutional neural networks

Abstract

The invention belongs to the field of image processing, aims at restoring a high-resolution depth image and utilizing the great learning capacity of convolutional neural networks to solve the defects that the conventional algorithm is high in computational complexity and high in actual application cost and cannot effectively extract features, and provides the technical scheme of a depth map super-resolution reconstruction method based on the convolutional neural networks. The convolutional neural networks (CNN) combining a convolutional layer and a deconvolutional layer is utilized to extract the depth image features of low-resolution sample depth image block and a high-resolution sample depth image block, and then the nonlinear mapping relation between the depth image features is learnt so as to restore the high-resolution depth image. The depth map super-resolution reconstruction method based on the convolutional neural networks is mainly applied to the occasion of image processing.

Description

technical field [0001] The invention belongs to the field of image processing, and relates to the optimization of a convolutional neural network in deep learning, specifically, to a depth map super-resolution reconstruction network combined with a convolutional layer and a deconvolutional layer. Background technique [0002] The acquisition of scene depth information is the focus of research in the field of computer vision. Depth information has important applications in 3D reconstruction, robot navigation, gesture recognition, movie games, and virtual scene modeling. In the early days, the stereo matching algorithm was used to obtain the disparity between the left and right images to obtain the depth information [1], but the algorithm has a certain limitation in the matching effect of the occluded area, the weak texture area, and the repeated texture area. Therefore, methods to directly obtain depth information have attracted people's attention. TOF cameras use a new type ...

AI Technical Summary

Problems solved by technology

[0013] In order to overcome the deficiencies of the existing technology, the present invention aims to restore high-resolution depth images, and utilize the huge learning ability of convolutional neural network to solve the shortcomings of traditional algorithms such as high computational complexity, inability to effectively extract features, and high practical application cost

Images