Form feature extraction method of chord length position matrix

A technology of shape features and extraction methods, which is applied in the field of image processing and can solve problems such as affine transformation is not robust

Inactive Publication Date: 2011-05-25
SHANDONG UNIV
View PDF2 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] Aiming at the problem that the existing chord-length correlation statistical shape feature extraction method is not robust to affine transformation, the present invention provides an anti- Affine Transformation Chord Length Position Matrix Shape Feature Extraction Method

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
  • Form feature extraction method of chord length position matrix
  • Form feature extraction method of chord length position matrix
  • Form feature extraction method of chord length position matrix

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] Applying the chord length position matrix shape feature extraction method of the present invention, the concrete steps of carrying out shape feature extraction to an image are:

[0025] 1. The object image is formed into a binary image by setting a threshold, that is, the object shape part is 1, and the rest is 0, and the width W and height H of the rectangular area where the object shape is located are calculated.

[0026] 2. Perform θ on the binarized image of the object shape i degrees of rotation, θ i ∈[0, 180), count the position of the chord length of the object in the vertical direction after rotation, and record the vertical direction as θ i degree direction. When counting the chord length, take N chords in each direction, then every two adjacent parallel chords c i，n-1 with c i，n And the area S of the region composed of the boundary n set as n ∈ [1, N]. if θ i Two adjacent parallel chords in the direction c i，n-1 with c i，n The interval between is Δρ...

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 form feature extraction method of a chord length position matrix, which comprises the following steps of: firstly, segmenting the shape of an object in each image from the image to form a binary image; secondly, rotating the shape of the object at an angle of theta i, 0<=theta i<180 degrees, and counting the position of a chord length of the object in the vertical direction after the rotation; thirdly, after determining the position of the chord length of the object in the vertical direction, counting the information of all the chord lengths in the position; and fourthly, counting the chord length in every directions and the features of position information to form a chord length position feature matrix, adopting a relative chord length and carrying out the normalization processing for the chord lengths in the chord length feature matrix, wherein the relative chord length is a ratio of an absolute chord length to a maximum chord length in the matrix. In the invention, the chord length position distribution information and the chord length ordering information with affine invariant are utilized, so that the recall ratio and the precision ratio rate of a system are ensured, and simultaneously, the commonality of similar objects can be better extracted, and the extracted shape features have the affine invariance and can resist affine transformation.

Description

technical field [0001] The invention relates to a shape feature extraction method in a content-based image retrieval system, belonging to the technical field of image processing. Background technique [0002] With the rapid development of multimedia technology, computer technology and Internet technology, images have become an important information resource. With the huge growth of people's demand for image information resources, image retrieval technology based on object shape has become a very important research direction of content-based image retrieval technology. Feature extraction and similarity measurement of target image are important research contents of image retrieval technology based on target shape. Based on the chord length correlation statistical algorithm, the chord length statistical histogram in each direction can be used to form a chord length correlation matrix, and the shape feature can be represented by this matrix. Experiments show that this feature ...

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 & AuthorityApplications(China)
IPC IPC(8): G06T7/00G06K9/46
Inventor杨明强柴华
OwnerSHANDONG UNIV