A Binocular Stereo Matching Method Based on Detail Enhancement

A binocular stereo matching and detailed technology, which is applied in image enhancement, image analysis, instruments, etc., can solve the problem of low matching accuracy and achieve the effect of reducing complexity and impact

Active Publication Date: 2021-11-02
WUHAN UNIV
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0008] In view of this, the present invention provides a binocular stereo matching method based on detail enhancement to solve or at least partly solve the technical problem of low matching accuracy existing in the methods in the prior art

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
  • A Binocular Stereo Matching Method Based on Detail Enhancement
  • A Binocular Stereo Matching Method Based on Detail Enhancement
  • A Binocular Stereo Matching Method Based on Detail Enhancement

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0041] The purpose of the present invention is to provide a high-precision binocular stereo matching method based on detail enhancement to solve the technical problem of low matching accuracy existing in the methods in the prior art. This method performs supervised learning and training on the stereo image pair data that has been marked with real disparity, and finally obtains a network that can perform stereo matching stably, and outputs a high-precision dense disparity map with rich details between two stereo images , so as to achieve the effect of improving the matching accuracy.

[0042] In order to achieve the above-mentioned technical effect, the inventive concept of the present invention is as follows:

[0043] First, use the disparity initialization sub-network to obtain the initial disparity estimation results at low resolution; then, through the optimization sub-network with a guided module designed in the present invention, combined with a multi-scale optimization s...

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 a binocular stereo matching method based on detail enhancement. First, the disparity initialization sub-network is used to obtain the initial disparity estimation result at low resolution; then, the optimization sub-network with a guiding module is combined with multi-scale optimization. strategy, gradually upsampling and optimizing the low-resolution initial disparity, and output the dense disparity estimation result at full resolution; then, using the multi-scale loss function, the disparity estimation result at low resolution is used as intermediate supervision to improve the network. Convergence accuracy to promote the disparity estimation results at full resolution; then input the binocular stereo image pair to be matched into the trained network to obtain the disparity estimation results. By learning and training the training data set with real disparity, this method not only greatly improves the accuracy and robustness of the disparity results obtained by stereo matching, but also significantly improves the accuracy of small objects and edges in the disparity map. Estimated results for detailed information.

Description

technical field [0001] The invention relates to the technical field of stereo matching in image processing, in particular to a detail enhancement-based binocular stereo matching method. Background technique [0002] Stereo matching is a classic problem in the field of image processing. Its main goal is to estimate the disparity (the column coordinates). The disparity map obtained by stereo matching has a wide range of applications in many fields, such as autonomous driving, indoor positioning, and 3D reconstruction. Therefore, stereo matching has important research value. [0003] In most traditional methods, stereo matching is usually divided into four steps, namely: calculating matching cost, cost aggregation, disparity estimation and disparity optimization. However, the matching cost calculated by traditional hand-designed features is often not robust in the face of challenging complex scenes, and thus limits the performance of traditional stereo matching methods. [...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/62G06N3/04G06T7/80
CPCG06T7/85G06T2207/20228G06N3/045G06F18/22G06F18/214
Inventor 姚剑谈彬陈凯涂静敏
Owner WUHAN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap