A single-view three-dimensional flame reconstruction method based on deep learning and a thin plate spline

A thin-plate spline and deep learning technology, applied in the field of computer vision, can solve the problems of increasing human, material and financial costs

Active Publication Date: 2019-06-21
ZHONGBEI UNIV
View PDF4 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem that multi-view images are required in the existing 3D flame reconstruction, which increases the cost of human, material and financial resources, the present invention proposes a single-view image based on deep learning and thin-plate splines. Figure three Dimensional Flame Reconstruction Method

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 single-view three-dimensional flame reconstruction method based on deep learning and a thin plate spline
  • A single-view three-dimensional flame reconstruction method based on deep learning and a thin plate spline
  • A single-view three-dimensional flame reconstruction method based on deep learning and a thin plate spline

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0057] The present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments.

[0058] Such as figure 1 with 2 As shown, the single vision based on deep learning and thin plate spline in this embodiment Figure three Dimension flame rebuilding method, it comprises the following steps:

[0059] Step 1, collecting data of different types of flames, constructing a two-dimensional image dataset and a three-dimensional model dataset of the flame;

[0060] This step is to collect two-dimensional images and three-dimensional models of many different types of flames on the Internet, wherein the types of flames include candle fire, gun fire and stove fire.

[0061] Step 2, using the method of deep learning to automatically find the precise features of the two-dimensional image of the flame and the three-dimensional model of the flame and classify the input flame;

[0062] This step extracts the two-dimensional image of the fla...

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 particularly relates to a single-view three-dimensional flame reconstruction method based on deep learning and a thin plate spline, which solves the problems that in the existing three-dimensional flame reconstruction process, multi-view images are needed to complete estimation and calculation of object depth information, the manpower and financial cost is high, and the like. The method comprises the following steps: firstly, searching and retrieving a three-dimensional flame model most similar to input flame in an existing data set through a deep learning method; Then, comparingwith the multi-angle projection view of the three-dimensional flame model to obtain an optimal projection view; And finally, processing the three-dimensional flame model by using a three-dimensionalthin plate spline deformation method to realize the reconstruction of the three-dimensional flame model. The method is suitable for three-dimensional reconstruction based on a single flame image or asingle-view flame image. The method is a relatively stable and relatively accurate three-dimensional model retrieval method; Compared with the mode that a control box is constructed on the basis of the outline to drive deformation, the time complexity is lower when the same control point is selected.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a single vision system based on deep learning and thin plate splines. Figure three Dimensional Flame Reconstruction Method. This method is suitable for 3D reconstruction based on a single flame image or a single-view flame image. Background technique [0002] Vision is an important means for human beings to perceive and understand the world. Computer vision technology allows computers to acquire, process, analyze and identify images by simulating human vision to realize the understanding of the real world. Three-dimensional reconstruction refers to a technology that restores the shape and spatial information of the three-dimensional object reflected from one or more images collected. Through 3D reconstruction, the lost 3D information of the detected object can be recovered and a complete 3D model can be constructed. [0003] The three-dimensional structure...

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): G06T17/00G06T7/13
Inventor 何利明焦世超陈佳瑜张建华
Owner ZHONGBEI 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