Image gradient domain local sequence coding and discrimination characteristic expression method

An image gradient and sequential encoding technology, applied in image encoding, image data processing, instruments, etc., can solve the problem of not fully considering the integration effect of gradient modulus length on recognition performance, not discussing edge misalignment and feature matching discontinuity, etc. question

Inactive Publication Date: 2016-03-23
SUN YAT SEN UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this method does not discuss the situation of edge misalignment and feature matching discontinuit

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
  • Image gradient domain local sequence coding and discrimination characteristic expression method
  • Image gradient domain local sequence coding and discrimination characteristic expression method
  • Image gradient domain local sequence coding and discrimination characteristic expression method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0046] The present invention will be further described below in conjunction with the drawings, but the embodiments of the present invention are not limited to this.

[0047] The local sequential encoding and discriminative feature representation method based on the image gradient domain of the present invention specifically includes:

[0048] 1. Image gradient features

[0049] Suppose the pixel value of each point of the image has a multiplicative structure Ω(x,y)=A(x,y)×L(x,y), where Ω(x,y) represents the image pixel value, A(x,y) ) Is the surface reflectance, L(x,y) is the illuminance at each point (x,y), the gradient feature of the image can be expressed as:

[0050] Ω x =(A×L) x ≈A x ×L+A×L x ,Ω y =(A×L) y ≈A y ×L+A×L y .(1)

[0051] According to the optical Albedo principle, it can be assumed that L changes very slowly, that is, L x ≈0,L y ≈0, the above equation can be further approximated as: Ω x =A x ×L,Ω y =A y ×L. Therefore, Ω a =Ω y. / Ω x ≈A y. / A x It can be roughly regard...

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 the computer vision and image mode analysis field and particularly relates to an image gradient domain local sequence coding and discrimination characteristic expression method. A local sequence coding scheme with image gradient direction constraint is firstly utilized to generate image texture characteristics, local difference of adjacent pixel points is clearly considered in image gradient filtering, the local texture structure coding capability can be enhanced, so intrinsic structure of a personal image can be effectively disclosed and expressed. According to the method, multi-mode characteristics are further automatically fused with a multi-core discrimination subspace, similarity measurement robustness can be effectively enhanced by utilizing the adaptive interaction function, and thereby inhibition on abnormal values generated in a characteristic matching process can be realized.

Description

Technical field [0001] The present invention relates to the field of computer vision and image pattern analysis, in particular to a gradient domain local sequential encoding and discriminative feature representation method of an image. It is an image encoding, non-linear dimensionality reduction, metric learning and discriminative representation method of portrait features under complex conditions . Background technique [0002] As a typical application scenario of supervised learning methods, face recognition has received extensive attention in public safety and commercial development. A good feature representation model plays a key role in robust face recognition. Although many algorithms have been proposed to deal with the effectiveness of feature design and extraction, many existing methods are still very sensitive to imaging conditions, including outdoor light, exaggerated expressions, and partial continuous occlusion. [0003] Representation-based subspace learning is one o...

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): G06T9/00
CPCG06T9/00G06T9/008
Inventor 任传贤戴道清
Owner SUN YAT SEN 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