Light field multi-view image super-resolution reconstruction method based on deep learning

A technology of super-resolution reconstruction and deep learning, applied in the field of light-field multi-view image super-resolution reconstruction based on deep learning, can solve the problem that the light-field multi-view image super-resolution method cannot meet the technical indicators, and achieve enhanced accuracy. Effect
CN112750076AActive Publication Date: 2021-05-04VOMMA (SHANGHAI) TECH CO LTD
4 Cites 9 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
VOMMA (SHANGHAI) TECH CO LTD
Publication Date
2021-05-04

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention discloses a light field multi-view image super-resolution reconstruction method based on deep learning. The method comprises the following steps: constructing a training set of high-resolution and low-resolution image pairs by adopting multi-view images which are obtained from a light field camera or a light field camera array and are distributed in an N * N array shape; constructing a multi-layer feature extraction network from the N * N light field multi-view image array to the N * N light field multi-view feature image; stacking the feature images and constructing a feature fusion and enhanced multi-layer convolutional network to obtain 4D light field structure features capable of being used for reconstructing a light field multi-view image; constructing an up-sampling module to obtain a nonlinear mapping relation from 4D light field structure features to a high-resolution N * N light field multi-view image; constructing a loss function based on the multi-scale feature fusion network, training the loss function, and finely adjusting network parameters; and inputting a low-resolution N * N light field multi-view image into the trained network to obtain a high-resolution N * N light field multi-view image.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The present invention relates to the technical field of image processing, in particular to a method for super-resolution reconstruction of light field multi-view images based on deep learning. Background technique

[0002] Light field cameras can simultaneously capture the spatial position and incident angle of light. However, the light field recorded by it has a trade-off relationship between spatial resolution and angular resolution. The limited spatial resolution of multi-view images limits the light field to a certain extent. The scope of application of the camera. The camera array is also restricted by cost and resolution, which limits the development of 3D light field display, 3D modeling, 3D measurement and other fields. With the continuous development of the field of image processing, the demand for super-resolution technology of light field multi-view images needs to be met urgently.

[0003] In recent years, the development and progress of ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More