Speckle image pixel-by-pixel matching method based on deep learning

A technology of speckle image and matching method, which is applied in the field of optical measurement, can solve the problems of large amount of calculation and achieve high precision

Active Publication Date: 2020-08-21
南京理工大学智能计算成像研究院有限公司
View PDF5 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, these matching algorithms all use multiple speckle images for three-dimensional measurement, and the calculation load is relatively large.
Therefore, there is still a lack of a single-frame speckle image matching method with high precision, high detail fidelity, and strong robustness.

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
  • Speckle image pixel-by-pixel matching method based on deep learning
  • Speckle image pixel-by-pixel matching method based on deep learning
  • Speckle image pixel-by-pixel matching method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0028] In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application, not to limit the present application.

[0029] Such as figure 1 , 2 As shown, this embodiment is a pixel-by-pixel matching method for speckle images based on deep learning, which can analyze a single frame of speckle images and obtain high-precision three-dimensional information. This method includes three steps.

[0030] step one. Collect training data and calculate the disparity value of the training data.

[0031] The binocular stereo vision system is used to project fringe and speckle patterns, and the left and right perspective cameras collect s different scenes. For each scene, a number of N phase-shifted fringe images...

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 speckle image pixel-by-pixel matching method based on deep learning, and the method can achieve the analysis of a single-frame speckle image, and obtains high-precision three-dimensional information. The method comprises the following steps: constructing a model based on a twin convolutional neural network, training the model through training data collected by a binocularstereoscopic vision system, obtaining parallax values of speckle images at left and right visual angles through the trained model, and combining the parallax values with calibration data of the binocular stereoscopic vision system to obtain three-dimensional information of an object. Compared with a method for calculating parallax through fringe projection, the method has the advantages that parallax data can be calculated only through one speckle image. Compared with a traditional speckle matching algorithm, the method has the advantages that the precision of a measurement result obtained bythe method is higher; compared with a basic convolutional neural network, the three-dimensional data reconstructed by the method has a higher-precision measurement result at details.

Description

technical field [0001] The invention relates to the technical field of optical measurement, in particular to a pixel-by-pixel matching method of speckle images based on deep learning. Background technique [0002] In recent years, 3D shape measurement technology has played an important role in the fields of medical imaging, robot navigation, face recognition, industrial inspection and human-computer interaction. Optical 3D measurement technology can be divided into passive 3D measurement and active 3D measurement according to the lighting method. Passive 3D measurement technology is to directly collect the 2D image of the object through the camera system, and determine its depth information, and finally form the 3D surface data of the object. Because this type of method generally has low measurement accuracy, it is inconvenient to use in fields such as face recognition and industrial inspection. Active 3D measurement technology projects images encoded according to certain ...

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 Applications(China)
IPC IPC(8): G06K9/62G06N3/04G06N3/08G06T5/50G06T7/33G06T7/80H04N13/239G01B11/25
CPCG06T7/337G06T7/85G01B11/25G06N3/08G06T5/50H04N13/239G06T2207/20081G06T2207/20084G06T2207/10012G06V10/751G06N3/045G06F18/22G06F18/214
Inventor 张晓磊左超沈德同
Owner 南京理工大学智能计算成像研究院有限公司
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