Method for extracting shape features of statistics correlated with relative chord lengths

A feature extraction and statistical shape technology, applied in computing, computer components, image analysis, etc., can solve the problem that affine transformation is not robust

Inactive Publication Date: 2012-01-04
SHANDONG UNIV
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] The present invention aims at the problem that the existing chord-length associated statistical shape feature extraction method is not robust to affine transformation, and uses the ratio of the chord length in each direction to the chord length passing through the centroid in this direction (that is, the relative chord length) Provides an anti-affine transform-resistant relative chord-length correlation statistical 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
  • Method for extracting shape features of statistics correlated with relative chord lengths
  • Method for extracting shape features of statistics correlated with relative chord lengths
  • Method for extracting shape features of statistics correlated with relative chord lengths

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0013] Applying the relative chord length correlation statistical shape feature extraction method of the present invention, the concrete steps of carrying out shape feature extraction to an image are:

[0014] 1. The image is formed into a binary image by setting a threshold, that is, part of the object shape is 1, and the rest is 0, and the width and height of the rectangular area where the object shape is located are calculated. Calculate the ratio t of the larger value L=max(W, H) to 128 in the width W and height H of the rectangular area where the shape of the object is located, and t=128 / L, and use the nearest neighbor method to zoom the image by t times.

[0015] 2. Find θ i degree and θ i Chord length histogram in the +90 degree direction. Transform the shape of the object into θ i degrees of rotation, θ i ∈[0, 90), after rotation, the statistics of the object in the vertical and horizontal directions (denoted as θ i degree and θ i +90 degree direction) the number...

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 method for extracting the shape features of statistics correlated with relative chord lengths. The method comprises the following steps: (1) segmenting the shape of an object from each image; (2) evaluating the histogram of the chord lengths in the direction of thetai degrees and thetai+90 degrees, rotating the shape of the object for thetai degrees, counting the number of the different chord lengths in vertical and horizontal directions of the positions of the rotated object on which the object is rotated for thetai degrees, and sequentially expressing the chord lengths in the direction of the positions covering an angle of 0 to 180 degrees in the histogram; (3) calculating the barycentric coordinates of the shape of the object and counting the chord lengths of the object passing through the barycenter of the object in the direction of 0 to 180 degrees, so as to obtain the directional principal chord; and (4) processing the array related to the chord lengths by using the directional principal chord and expressing the shape features by a normalized array. The invention can guarantee the recall ratio and precision ratio of the system, better extract the similarity of similar objects and ensure that the extracted shape features have affine invariance.

Description

technical field [0001] The invention relates to a shape feature extraction method for a content-based image retrieval system and an automatic statistical system for similar objects. Background technique [0002] With the rapid development of network technology, information retrieval tools provide a convenient and efficient method for the majority of netizens to obtain the information they need. The current mature information retrieval tools use text as the search term to search for relevant information in the network, but content-based image retrieval technology is still immature, one of the important reasons is the difference between computer vision and human perception. In the image shape feature extraction algorithm, the similarity feature is an important content. Similarity features refer to the common features of the same type of objects. As a computer vision, it is a difficult task to extract the common features of similar but not identical objects. Based on the sta...

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