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

Photometric stereo vision data driving global optimization algorithm for solving BRDF high nonlinearity problem

A photometric three-dimensional, data-driven technology, applied in image data processing, computing, instruments, etc., can solve the problems of cumbersome optimization iterative solution process, slow calculation speed, complex model, etc., and achieve the effect of ensuring training accuracy and prediction speed

Pending Publication Date: 2019-09-17
舒轶
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the adaptability of the above models is still limited, and there is still room for improvement in the accuracy of normal vector recovery
In addition, the proposed models are often relatively complex, or the optimization iterative solution process is cumbersome and the calculation speed is slow

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
  • Photometric stereo vision data driving global optimization algorithm for solving BRDF high nonlinearity problem
  • Photometric stereo vision data driving global optimization algorithm for solving BRDF high nonlinearity problem
  • Photometric stereo vision data driving global optimization algorithm for solving BRDF high nonlinearity problem

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0033] The present invention is further described below in conjunction with accompanying drawing:

[0034] In general, the BRDF of non-Lambertian materials is highly nonlinear, and the BRDF of different materials is not the same. Therefore, it is more appropriate to use the dense measurement data of a certain material to represent the BRDF of the material than the unified analytical model. The basic idea of ​​the present invention is to obtain a reasonable mathematical model according to the measurement database of a certain material, and then infer the data in the corresponding material database according to the data model. Specifically, the present invention establishes the incident direction l, the observation direction v, the pixel record value I p The mapping model between BRDF and BRDF, and perform data inference according to the mapping model.

[0035] In order to realize the above idea, the present invention uses Gaussian process to establish a mathematical model. A...

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 relates to a photometric stereo vision data driving global optimization algorithm for solving a BRDF high nonlinearity problem in the technical field of machine vision identification, and provides a mapping relation for describing a non-Lambert reflection phenomenon, and the mapping relation realizes decoupling of the reflection phenomenon and a surface normal vector while describing the reflection phenomenon. Aiming at the proposed mapping relationship, a corresponding continuous mathematical model is established by using a Gaussian process, and a sampling strategy is designed to ensure the training precision and prediction speed of the model. According to the method, the problem of high nonlinearity of the BRDF can be better solved, and the calculation efficiency is relatively high. The simulation experiment and the real experiment based on the MERL database verify the superiority of the method. The method has a good application prospect in the aspect of defect detection of a large batch of single materials.

Description

technical field [0001] The invention relates to a technology in the field of machine vision recognition, in particular to a photometric stereo vision data-driven global optimization (Data-driven Photometric Stereo, DPS) algorithm that solves the highly nonlinear problem of BRDF. Background technique [0002] One of the difficulties in photometric stereo vision technology is to quickly and effectively solve the complex reflection problem of non-Lambertian materials. Traditional photometric stereo vision estimates object surface normal vectors based on the Lambertian body assumption. Under this assumption, there is a linear relationship between the pixel record value and the object surface normal vector. Therefore, for a certain pixel, if the pixel record values ​​of three non-coplanar illumination directions are known, the normal vector of the pixel corresponding to the surface point of the object can be calculated. But in the real world, there are few materials satisfying ...

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): G06T7/00G06T5/00
CPCG06T7/0002G06T2207/20032G06T5/90
Inventor 舒轶任明俊
Owner 舒轶
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