Pattern matching recognition system and implementing method thereof

A recognition system and implementation method technology, applied in character and pattern recognition, image analysis, image data processing and other directions, can solve the problems of lower recognition accuracy, limited number of simple shape marks, etc., to avoid repeated matching confirmation, The effect of improving matching efficiency

Inactive Publication Date: 2009-02-04
TONGJI UNIV
View PDF0 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Due to the large number of feature points, if only color features are used, the recognition a

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
  • Pattern matching recognition system and implementing method thereof
  • Pattern matching recognition system and implementing method thereof
  • Pattern matching recognition system and implementing method thereof

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0051] (1) Content of training samples

[0052] In this embodiment, 8 shapes of artificial marker points with 3-6 convex corners are selected, such as image 3 shown, and take red (R255G0B0), green (R0G255B0), blue (R0G0B255), yellow (R255G255B0), purple (R255G0B255), blue (R0G255B255), white (R255G255B255), black ( R0G0B0), be the test color, and color feature and shape feature are as shown in table 1. Take more corner points and more color types, and there will be more types of identification points. In this example, 8*8=64 kinds of marking points can be generated, which can satisfy the number of manual marking points required for 3D reconstruction of medium-sized objects.

[0053]

[0054] Table 1 Characteristic vector data

[0055] (2) Format of training samples

[0056] Define a sample as (R', G', B', γ', β', X, X, X, X, X, X). The first 5 bits are input parameters, which are normalized color tristimulus value vector and normalized shape vector...

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 shape matching identification system and the implementation method thereof, aiming at the problems in the prior art after the color and shape identifications are researched. Firstly, a digital camera is adopted to shoot the object adhered with an artificial label for the three-dimensional reconstruction and form the picture of the object; then the shape characteristic and the color characteristic of the picture are extracted through the characteristic extraction module; the shape characteristic and the color characteristic are written into a characteristic vector and normalized through the characteristic vector normalization module; finally, the normalized characteristic vector is recognized by a neural network module. The invention has the advantages that the system and the implementation method adopt the mixed matching method based on color and shape to avoid the gray-scale based matching and the multiple repeated matching acknowledgements with large computational complexity; based on the Cartesian product combination of the angle point number characteristic in the color category and the space characteristic, the artificial label point characteristics with enough quantity are generated to enable the matching operation to be finished in one-step, so as to improve the matching efficiency greatly.

Description

【Technical field】 [0001] The invention relates to a recognition system and its realization method, in particular to a pattern matching recognition system and its realization method. 【Background technique】 [0002] With the development of technology, matching has always played an important role in the 3D reconstruction process based on camera calibration. In the process of 3D reconstruction of large scene objects, in order to facilitate the extraction of object feature points and reduce the difficulty, we will artificially paste some calibration points on the object. Due to the relatively large scene, it is impossible to ensure that the target at a fixed position is captured every time during the shooting process of the camera. To solve this difficulty, these calibration points are divided into two types according to size. The large points are called skeleton points, and the small points are called is an unknown point. First calculate the three-dimensional world coordinates...

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): G06K9/00G06K9/62G06T7/40
Inventor 郝泳涛
Owner TONGJI UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap