Target detection method based on improved HOG-ULBP characteristic operator

A feature operator and target detection technology, applied in computing, computer parts, instruments, etc., can solve the problems of large increase in feature vector dimension, high real-time requirements, and poor real-time performance, so as to reduce detection time and realize real-time performance. The effect of enhancing sexuality and improving the detection rate

Inactive Publication Date: 2017-03-29
SOUTH CHINA UNIV OF TECH
View PDF4 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, this method has the following disadvantages: 1) The dimension of the feature vector increases too much after feature merging, which easily leads to over-fitting phenomenon; Real-time deterioration

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
  • Target detection method based on improved HOG-ULBP characteristic operator
  • Target detection method based on improved HOG-ULBP characteristic operator
  • Target detection method based on improved HOG-ULBP characteristic operator

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0047] Hereinafter, the present invention will be further described in detail with reference to the examples and drawings, but the implementation of the present invention is not limited thereto.

[0048] In this embodiment, a target detection method based on an improved HOG-ULBP feature operator, such as figure 1 , Including the following steps:

[0049] The learning phase includes:

[0050] S1.1: Establish a positive and negative sample library for training.

[0051] The selection of the positive and negative sample bank follows the following two rules:

[0052] Rule 1: The ratio of negative samples to positive samples is 10:1;

[0053] Rule 2: Use the first training SVM to detect false negative samples in negative samples as difficult negative samples, and increase the proportion of these difficult negative samples. Thereby, the accuracy of the established model can be further improved.

[0054] S1.2: Extract the region of interest from the sample.

[0055] It can be manually extracted...

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 target detection method based on an improved HOG-ULBP characteristic operator. The method comprises a learning stage and a decision stage, for the learning stage, a positive and negative sample database is established; interest regions are extracted for samples; an HOG characteristic and a ULBP characteristic are extracted, the HOG characteristic and the ULBP characteristic are combined to form an HOG-ULBP vector, and then Gaussian normalization processing is carried out; a local coordinate code (LCC) is utilized to acquire an improved HOG-ULBP characteristic operator; a linear SVM model is established according to the improved HOG-ULBP characteristic operator; for the decision stage, an HOG characteristic and a ULBP characteristic are extracted from a to-be-detected video frame image, a linear SVM model is acquired according to the improved HOG-ULBP characteristic operator and is then outputted, if model output is determined to be a positive sample, a target is detected, and the target position is outputted. The method is advantaged in that an over-fitting phenomenon caused by over-large dimension after combination of the HOG characteristic and the ULBP characteristic is effectively solved, and a target object detection rate is improved.

Description

Technical field [0001] The invention belongs to the field of computer vision target detection, and particularly relates to a target detection method based on an improved HOG-ULBP feature operator. Background technique [0002] With the rapid development of computer technology, communication technology and network technology, contemporary society has been in an era of information explosion, and our lives are full of all kinds of information. Especially with the rise of the Internet, images and videos have increasingly become the main forms of carrying information. How to process massive amounts of image and video information is a hot and difficult issue to be solved urgently. [0003] Target detection is one of the research hotspots in the field of image processing. Target detection has a wide range of applications in real life, such as road detection in aircraft aerial photography or satellite images, vehicle and pedestrian detection in video surveillance, tumor detection in CT i...

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): G06K9/62G06K9/46
CPCG06V10/50G06F18/2411
Inventor 冯颖杨涛郑佳泰杜娟陈新开
Owner SOUTH CHINA UNIV OF TECH
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