Unlock instant, AI-driven research and patent intelligence for your innovation.

3D Model Retrieval Method Based on Image Edge and Line Features

A 3D model and image feature technology, applied in still image data retrieval, still image data indexing, network data retrieval, etc., can solve problems such as difficulty in viewpoint selection and redundancy of 3D model views, so as to solve the difficulty in viewpoint selection and improve retrieval efficiency effect

Active Publication Date: 2019-05-24
ZHEJIANG UNIV OF TECH
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Solved the problem of redundancy and difficult viewpoint selection in 3D model view extraction

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
  • 3D Model Retrieval Method Based on Image Edge and Line Features
  • 3D Model Retrieval Method Based on Image Edge and Line Features
  • 3D Model Retrieval Method Based on Image Edge and Line Features

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0047] Example: such as figure 1 As shown, the 3D model retrieval method based on image edge and line features includes the following steps

[0048] Step 1: Firstly, crawl the 3D model and a single context image in an intuitive single correspondence manner to form a 3D model and its context image set.

[0049] Step two, such as figure 2 As shown, color moments are used to initially perform coarse clustering for model context images. Similar to the color histogram, the color moment is a more effective representation of color features. Using the concept of moments in linear algebra, the color distribution in the image is represented by its moments. Use the color first-order moment (mean value), color second-order moment (variance) and color third-order moment (skewness) to describe the color distribution, we have

[0050]

[0051] Among them, p i,j is the occurrence probability of the pixel with the gray value j in the i-th color channel of the image, and N is the total ...

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 an image edge and line feature based three-dimensional model retrieval method. The method comprises the steps of firstly, establishing an original database by crawling three-dimensional model data and a context image released in the internet; secondly, by extracting a color feature of the context image and a shape feature of the three-dimensional model, performing clustering preprocessing on the original database and defining a primary category weight; and based on this, extracting an edge line as well as a local Zernike moment feature and a Gabor filtering feature for the image, and forming a context image original feature library in a weighting form. Likewise, by adopting a bag-of-visual-words method, the context image subjected to similarity sorting is retrieved from the feature library after encoding and weight computing, so that the three-dimensional model retrieval is realized; and the problems of view redundancy and difficulty in view point selection are solved.

Description

technical field [0001] The invention relates to the field of computer graphics, in particular to a three-dimensional model retrieval method based on image edge and line features. Background technique [0002] In the field of 3D model retrieval, image is the most simple and natural retrieval interface. The core problem of 3D model retrieval using images is how to evaluate the similarity between images and 3D models. Theoretically speaking, the search results can be obtained by similarity matching between the shape of the image and the 3D model view, but there are two problems in this research idea: [0003] (1) Multiple views are required for any 3D model to ensure retrieval accuracy, which greatly reduces retrieval efficiency. Each view of the 3D model represents the posture of the same 3D model under different viewpoints. In the existing methods, tens or even hundreds of uniform views are usually extracted for the 3D model to ensure a sufficiently comprehensive retrieval...

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): G06F16/583G06F16/53G06F16/51G06F16/953
CPCG06F16/51G06F16/5838G06F16/951
Inventor 刘志尹世超潘翔
Owner ZHEJIANG UNIV OF TECH