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

Method for optimizing depth image in RGBD sequence scene flow calculation

A technology of RGB image and depth image, applied in the field of depth image optimization in RGBD sequence scene stream calculation, which can solve problems such as limitations

Active Publication Date: 2017-07-14
NANCHANG HANGKONG UNIVERSITY
View PDF3 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The existing scene flow calculation depth image layering method usually adopts artificially setting the number of layers, because the motion and scene information of different types of RGBD sequence images vary widely, so it is greatly limited in practical applications

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
  • Method for optimizing depth image in RGBD sequence scene flow calculation
  • Method for optimizing depth image in RGBD sequence scene flow calculation
  • Method for optimizing depth image in RGBD sequence scene flow calculation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] The present invention will be further described below in conjunction with drawings and embodiments. see Figure 1a to Figure 4 , the depth image optimization method in RGBD sequence scene flow calculation, using the bear_back image sequence to perform experiments on automatic layering and optimal segmentation of depth images:

[0025] 1) Since image scenes are usually decomposed into a small number of independent moving objects, such as Figure 1a and Figure 1b It is a bear_back image sequence with two consecutive frames of images (where: Figure 1a is the first frame image, Figure 1b is the second frame image), Figure 1c Yes Figure 1a Corresponding depth image; first set the initial layer number N=8;

[0026] 2) Calculate the optical flow between two consecutive frames of the bear_back image sequence, such as Figure 2a shown; and according to the initial layer number of the depth image corresponding to the bear_back image sequence K-means clustering to obtain ...

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 method for optimizing a depth image in RGBD sequence scene flow calculation. The method comprises steps of: firstly setting the number of initial segmentation layers and obtaining a depth image initial segmentation result by using a K-mean clustering algorithm; then calculating an RGB image sequence inter-frame optical flow and determining whether to merge adjacent layers by using an average angular error and a point error between the adjacent layers in the depth image initial segmentation result; and finally, cyclically determining the depth image initial segmentation result in order to obtain the number of automatic segmentation layers and the final segmentation result of the depth image in the RGBF sequence scene flow calculation when the number of depth image segmentation layers is no longer changed. Compared with a conventional depth image manual layer segmentation method, the method can realize the automatic layer segmentation of the depth image, and is more accurate in the segmentation result.

Description

technical field [0001] The invention relates to a depth image automatic layering method, in particular to a depth image optimization method in RGBD sequence scene flow calculation. Background technique [0002] Scene flow is a three-dimensional motion field formed by the movement of spatial scenes or objects. Scene flow expands the motion estimation of scenes or objects from two-dimensional to three-dimensional, and has important applications in robot vision, drone navigation, virtual reality, and remote control. value. Estimating scene flow with RGBD sequences has gained increasing attention as consumer-grade depth sensors become widely available. Although depth image information can restore the 3D motion and structure of a scene or object from a single-view RGB image, when the edge of the scene or object in the depth image does not accurately match the edge of the RGB image, the scene flow calculation model based on the RGBD sequence will fail. Determine the motion bound...

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): G06T7/10G06T7/579G06K9/62
CPCG06F18/23213
Inventor 陈震张聪炫朱令令何超江少锋
Owner NANCHANG HANGKONG UNIVERSITY
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