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

Multi-view target identification method based on view semantic information and sequence context information

A semantic information and target recognition technology, applied in the field of multi-view target recognition, can solve the problems of large amount of calculation, simple and rough structure, etc., and achieve the effect of improving accuracy

Inactive Publication Date: 2020-11-10
TIANJIN UNIV
View PDF5 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] The present invention provides a multi-view target recognition method based on view semantic information and sequence context information. Inspired by the successful application of 3D CNN in the field of video processing, the present invention introduces 3DCNN to design a new network structure for the mining of sequence context information. It overcomes the shortcomings of the existing methods in the study of sequence context, such as simple and rough structure or large amount of calculation, and on this basis, it adds attention to the view itself, uses the semantic information of the view to enhance the view features, and improves the multi-view target recognition. The accuracy of , see the description below:

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
  • Multi-view target identification method based on view semantic information and sequence context information
  • Multi-view target identification method based on view semantic information and sequence context information
  • Multi-view target identification method based on view semantic information and sequence context information

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035] A multi-view target recognition method based on view semantic information and sequence context information, see figure 1 , the method includes the following steps:

[0036] 101: Use a virtual camera to take virtual photos of the 3D model in the 3D model database, and generate a view sequence of the 3D model;

[0037] 102: Use a neural network to extract each view feature in the view sequence, and generate a view feature sequence;

[0038] 103: Learn the semantic information of each view feature in the view feature sequence, enhance the useful semantic information in the view feature, and curb the interference information;

[0039] In the existing technology, the view semantic information often only pays attention to the relationship between views, and rarely returns to the view itself, that is, the view semantic information. Therefore, this method pays attention to the characteristics of the view itself, making up for the blind spots of previous research.

[0040] 104...

Embodiment 2

[0045] The scheme in embodiment 1 is further introduced below in conjunction with specific examples and calculation formulas, see the following description for details:

[0046] 201: First, use the virtual camera to take virtual photos of the models in the 3D model database to generate a view sequence;

[0047] Wherein, the above step 201 mainly includes:

[0048] A set of viewpoints is predefined, and the viewpoints are the viewpoints for observing the target object. In the embodiment of the present invention, 12 viewpoints are set, that is, a virtual camera is placed every 30 degrees around the center of mass of the 3D model, and the viewpoints are evenly distributed on the target around the object. By selecting different interval angles, different angle views of the 3D model are acquired clockwise to generate a view sequence.

[0049] 202: Use a neural network to extract each view feature in the view sequence, and generate a view feature sequence;

[0050] Among them, th...

Embodiment 3

[0073] Below in conjunction with concrete test, the scheme in embodiment 1 and 2 is carried out feasibility verification, see the following description for details:

[0074] The present invention adopts the data set disclosed by ModelNet40, and compares it with other multi-view target recognition methods, and the evaluation indexes select classification accuracy and mAP respectively. [7] .

[0075]

[0076]

[0077] From the above experimental data, it can be seen that the multi-view target recognition method based on view semantic information and sequence context information proposed by the present invention has better performance than the current mainstream methods, and can well handle the challenges faced in multi-view target recognition.

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 multi-view target recognition method based on view semantic information and sequence context information, and the method comprises the steps: carrying out the virtual photographing of a three-dimensional model in a three-dimensional model database through a virtual camera, and generating a view sequence of the three-dimensional model; extracting each view feature in the view sequence by using a neural network to generate a view feature sequence; learning semantic information of each view feature in the view feature sequence, enhancing useful semantic information in the view features, and suppressing interference information; utilizing 3D CNN to learn relevance between adjacent view features in the enhanced view feature sequence, and mining context information of the view feature sequence; and finally, fusing the view feature sequences into a compact feature descriptor through global average pooling, and performing multi-view target identification by utilizingthe feature descriptor. According to the method, the semantic information of the view is used to enhance the view features, and the precision of multi-view target recognition is improved.

Description

technical field [0001] The invention relates to the field of view sequence and multi-view target recognition, in particular to a multi-view target recognition method based on view semantic information and sequence context information. Background technique [0002] In recent years, with the wide application of 3D technology in virtual reality, 3D printing, medical diagnosis and other fields [1] , the number of 3D objects is growing rapidly, making multi-view object recognition methods receive great attention. Meanwhile, a lot of work has been devoted to constructing discriminative descriptors [2] . existing methods [3] Usually, multiple views of a 3D object are obtained by placing a virtual camera around it, and then the features of each view are extracted by a neural network, and finally these view features are fused into a compact feature descriptor. On this basis, some well-known databases such as ModelNet40 are also derived. [4] , there are many researchers conductin...

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): G06K9/00G06K9/62G06T17/00G06N3/04G06N3/08
CPCG06T17/00G06N3/08G06V20/64G06V20/41G06V20/46G06N3/045G06F18/253
Inventor 刘安安郭富宾周河宇宋丹
Owner TIANJIN UNIV