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

A Hierarchical Image Segmentation Method Based on Multi-Scale Edge Clues

An image segmentation and multi-scale technology, applied in the field of image processing, can solve the problems of inability to guarantee the consistency of segmentation results, inability to utilize and control multi-scale image information, and achieve the effect of maintaining consistency, improving effect, and clear structure

Active Publication Date: 2022-03-15
NANJING UNIV
View PDF1 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

To sum up, there are two main problems in the existing hierarchical segmentation methods: first, the semantic consistency of the segmentation results of each layer cannot be guaranteed; second, the multi-scale image information cannot be reasonably rationalized in the segmentation process use and control

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 Hierarchical Image Segmentation Method Based on Multi-Scale Edge Clues
  • A Hierarchical Image Segmentation Method Based on Multi-Scale Edge Clues
  • A Hierarchical Image Segmentation Method Based on Multi-Scale Edge Clues

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0070] In this embodiment, such as Figure 2a Shown is the input image to be segmented, Figure 2b ~ 2f In order to scale the input image and separate the multi-scale image content components after the content style, this embodiment shows the results under 5 scales, which are respectively scaled to the original image 0.2, 0.4, 0.6, 0.8 and 1.0 in order. Figure 3b ~ 3f is an edge probability map for edge detection on image content components, with Figure 2b ~ 2f hint Figure 1 One to one correspondence. Figure 4a~4d is the generation process from an input image to a hierarchically merged tree, where Figure 4b for Figure 4a The superpixelated result of , Figure 4c for will Figure 4b The graph structure representation after being transformed into a region adjacency graph, Figure 4d for right Figure 4c Hierarchical merge tree generated after iterative region merge. Figure 5 Represents the process of optimizing a hierarchically merged tree through energy function...

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 hierarchical image segmentation method based on multi-scale edge cues, comprising the following steps: performing superpixelation on an input image to obtain an over-segmented region of the image; performing multi-scale content style separation on the input image, thereby extracting The content components of the image, and use the edge detection method to detect the edge information of the image content components, and obtain a multi-scale edge probability map; measure the similarity between the over-segmented regions according to the multi-scale edge clues, and perform iterative region merging to generate a hierarchy Merge tree: establish an energy function based on the hierarchical merge tree and prior information, and use the dynamic programming method to solve the energy function to optimize the hierarchical merge tree structure and realize the hierarchical segmentation of the image.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a hierarchical image segmentation method based on multi-scale edge clues. Background technique [0002] As a very important research work in the field of computer vision, image segmentation has been widely used in many image processing algorithms and applications, such as object detection, image retrieval and style transfer. The general image segmentation problem is defined as: "dividing an image into different regions whose pixels have similar characteristics". Hierarchical segmentation can be regarded as a combination of coarse-to-fine multi-layer image segmentation results, so that a single multi-scale structure can be used to segment objects at all scales in the image, and achieve multi-scale processing and representation of images. Compared with the traditional single-layer image segmentation, it can show more image information. The result of hierarchical segmenta...

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): G06T5/00G06T7/11G06T7/13G06T7/44G06T7/90G06V10/774G06K9/62
CPCG06T7/11G06T7/13G06T7/44G06T7/90G06F18/232G06T5/70
Inventor 孙正兴徐峻峰胡安琦王爽
Owner NANJING 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