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

Variable-scale image segmentation method based on RGB-D

A RGB-D, variable scale technology, applied in image analysis, image enhancement, image data processing and other directions, can solve problems such as increasing the amount of calculation, and achieve the effect of multi-scale analysis, strong robustness, and strong interpretability

Inactive Publication Date: 2019-12-27
NANJING UNIV OF SCI & TECH
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The currently commonly used supervoxel segmentation method has a high consistency in the size of each supervoxel obtained after the scale parameter is determined. If multi-scale analysis is required in the future, the size of the supervoxel can be controlled by setting different scale factors. When analyzing each scale, it is necessary to calculate the super-voxel segmentation result of a scale, which increases the amount of calculation

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
  • Variable-scale image segmentation method based on RGB-D
  • Variable-scale image segmentation method based on RGB-D
  • Variable-scale image segmentation method based on RGB-D

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0073] figure 1 It is the result map obtained after the processing of one frame of RGB-D image obtained by shooting one frame of mynt_eye in the present invention. The following method is used to segment the picture, including the following steps.

[0074] Step 1. Acquisition of the depth image gradient map. Set a frame of RGB-D image data as P, with a resolution of 480 rows and 640 columns. Each data point includes 3 color channels (R, G, B) and a depth channel ( d), use formula (1) to solve the gradient of the depth image, and obtain the gradient map D of the depth image.

[0075]

[0076] Step 2:

[0077] For the selection of seed points, randomly select a point p0 in D as the seed point, set the radius R as the minimum threshold distance of the seed point to be generated, and use Poisson castellation sampling algorithm to sample in D to obtain the set of seed points;

[0078] Step 2-1. Randomly select a seed point seed in the gradient map of the depth image D obtained...

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 provides a variable-scale image segmentation method based on RGB-D. The method comprises: performing Poisson sampling by using a gradient map corresponding to a depth map to select seedpoints, obtaining initial super-pixel points based on gradient change, then establishing a map model according to the adjacency relation of super-pixels, performing super-pixel fusion by using a map method, and finally obtaining a segmented final result. According to the method, the gradient information of the depth map is fully utilized, and the method is more convenient and quicker.

Description

technical field [0001] The invention belongs to image segmentation technology, in particular to a variable-scale image segmentation method based on RGB-D. Background technique [0002] With the advancement of sensor technology, the acquisition cost of RGB-D images is getting lower and lower. How to preprocess RGB-D images more effectively is an important research content of computer vision in recent years. In order to make full use of the three-dimensional geometric information in RGB-D images, similar to the concept of super-pixel over-segmentation in two-dimensional images, over-segmenting RGB-D images into super-voxels is an effective preprocessing method, which can effectively Reduce the amount of data processed by subsequent algorithms. The currently commonly used supervoxel segmentation method has a high consistency in the size of each supervoxel obtained after the scale parameter is determined. If multi-scale analysis is required in the future, the size of the superv...

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): G06T7/11G06T7/187
CPCG06T7/11G06T7/187G06T2207/10028
Inventor 钟易潘丁永良宋迪袁夏
Owner NANJING UNIV OF SCI & TECH
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