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

A Joint Inference Method for 3D Facade Layout Estimation Based on Day-Night Image Pairs

A technology of three-dimensional facade and image pairing, applied in the field of urban image understanding, to achieve the effect of accurate estimation results

Inactive Publication Date: 2020-04-28
BEIHANG UNIV +1
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this type of technology has very high requirements for image quality, and it needs to include at least a whole day or even a whole year of data to obtain accurate surface normal and depth

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 Joint Inference Method for 3D Facade Layout Estimation Based on Day-Night Image Pairs
  • A Joint Inference Method for 3D Facade Layout Estimation Based on Day-Night Image Pairs

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0035] The present invention will be described in further detail below in conjunction with the accompanying drawings.

[0036] Such as figure 1 As shown, the overall schematic diagram of the present invention, first based on a group of aligned day and night image pairs, detect the lattice structure of the night image and the boundary and geometric information of the daytime image; then map the lattice to the daytime image, based on the lattice pair image Carry out segmentation; in the process of segmentation, it is necessary to continuously update the boundary of the facade and the lattice structure of the core, and finally calculate the depth and orientation of the facade area to obtain the three-dimensional facade layout.

[0037] When obtaining the dot matrix structure of the night image, based on the night city environment, artificial light sources, such as window lights, etc. bring people visually easy to observe characteristics, using its local high contrast and regular...

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 joint inference method for carrying out 3D facade layout estimation based on day and night images, and geometric information in an extracted night image is used to assist the facade layout estimation of a daytime image. The method comprises the steps of (1) detecting the night image based on semantic window lamp information, extracting the geometric information of a 3D facade window from the image, and constructing a lattice structure of the night image, (2) carrying out semantic segmentation on the daytime image, and obtaining surface layout of a daytime image structure according to the context relation of an semantic object after segmentation, and (4) constantly updating the lattice structure based on an inference framework of iteration, thus calculating new facade boundary and direction, and obtaining new surface layout. Compared with the estimation of facade layout with the only use of the daytime image, more sufficient information is used, and an accurate result is obtained.

Description

technical field [0001] The invention relates to the field of urban image understanding, and more specifically, relates to a joint inference method for estimating three-dimensional facade layout based on day and night image pairs. Background technique [0002] A very classic technique in scene understanding is to estimate its 3D facade layout from a single image, but due to the lack of features that represent the quantitative properties of the image space, there has been no revolutionary progress in obtaining accurate 3D facade layout structures . Zeng et al. demonstrated in the field of image segmentation that the repeated structure of the middle layer can improve the accuracy of the 3D facade layout structure. Analyzing the daytime images found some debris, such as shadow changes on buildings and discontinuities in depth in the scene, which will affect the effect of repeated structure detection. After analysis, it is found that the night images under the same viewpoint co...

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): G06T15/00
CPCG06T15/005
Inventor 周忠周颐孟明庚琦川刘俊琦胡勇刘睿严柯吴威
Owner BEIHANG 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