Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

RGB image guided depth map super-resolution method based on joint implicit image function

A RGB image and super-resolution technology, applied in the field of RGB image-guided depth map super-resolution based on joint implicit image functions, can solve the problems of poor interpretability and achieve good interpretability

Pending Publication Date: 2021-12-28
PEKING UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

This method can take advantage of the feature extraction ability of CN

Method used

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
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • RGB image guided depth map super-resolution method based on joint implicit image function
  • RGB image guided depth map super-resolution method based on joint implicit image function
  • RGB image guided depth map super-resolution method based on joint implicit image function

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0022] The technical solutions in the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention.

[0023] The present invention proposes a depth map super-resolution image restoration method based on joint implicit image function guidance, figure 1 Shown is the flow of the RGB image-guided depth map super-resolution method provided by the present invention; by proposing a joint implicit image function (Joint Implicit Image Function, JIIF), from the perspective of implicit neural representation (Implicit Neural Representation) processing RGB Image-Guided Depth Map Super-Resolution Restoration Task. Specifically, we solve the super-resolution problem in the form of image interpolation, and use an MLP to learn the weights and values ​​used in the interpolation process, and finally calculate the predicted pixel values ​​​​through the interpolation formula. The method of the present invention is ap...

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

PUM

No PUM Login to View More

Abstract

The invention discloses an RGB image guided depth map super-resolution method based on a joint implicit image function, and the method comprises the steps: building a deep neural network model which is used for depth map super-resolution image restoration and comprises an RGB image encoder, a depth map encoder and a JIIF decoder; adopting an implicit neural representation method to establish a joint implicit image function suitable for an RGB-guided depth map super-resolution task, and carrying out modeling on multi-modal input; respectively extracting features from the RGB image and the input low-resolution depth map through two encoders of the established deep neural network model; predicting and outputting the depth value of each pixel under the high resolution through a JIIF decoder; and realizing the super-resolution image recovery of the depth map guided by the RGB image based on the combined implicit image function. According to the method, the image recovery effect superior to that in the prior art is achieved on the RGB image guided depth image super-resolution recovery task, and the method has better interpretability.

Description

technical field [0001] The invention relates to a RGB image-guided depth map super-resolution (image restoration) method based on a joint implicit image function, which can be applied to RGB image-guided depth map super-resolution tasks, and specifically relates to a joint implicit image function The definition of and a deep neural network model for learning this function from data are used to restore a low-resolution, noise-containing depth map to a high-resolution, noise-free depth map, belonging to the field of computer vision image processing technology. Background technique [0002] The RGB image-guided depth map super-resolution task refers to the task of recovering a high-resolution, noise-free depth map from a low-resolution, noisy depth map and a corresponding high-resolution RGB image, It has practical applications in tasks such as 3D reconstruction. Limited by the accuracy of the depth sensor, the quality of the captured depth map is usually low, but the developm...

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

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06T3/40G06T7/50G06T7/90
CPCG06T3/4053G06T7/50G06T7/90G06T3/4007
Inventor 唐嘉祥陈小康曾钢
Owner PEKING UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com