A sample expansion method fusing virtual samples

A technology of virtual samples and sample expansion, applied in the field of increasing the number of car logo samples, can solve the problems of small car logo data sets and inability to generate a large number of virtual samples, and achieve the effect of improving accuracy and adaptability

Active Publication Date: 2018-12-18
ZHEJIANG UNIV OF TECH
View PDF1 Cites 8 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Since the method of using prior knowledge to expand samples and the method of perturbation have certain limitations, a large number of virtual samples cannot be generated. Therefore, in view of the problem of small car logo data sets, the present invention proposes a sample expansion method that combines virtual samples. 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
  • A sample expansion method fusing virtual samples
  • A sample expansion method fusing virtual samples
  • A sample expansion method fusing virtual samples

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0032] The specific implementation method of the internal truck positioning method will be described in detail below in conjunction with the embodiments. The specific examples described here are only used to explain the present invention, not to limit the present invention.

[0033] The specific method of removing the vehicle logo is as follows:

[0034] Step 1: Mark the collected real vehicle image I, and mark the position as R(x, y, w, h), where the upper left corner of the image I is the origin of the pixel coordinates, and (x, y) is the rectangular area R The coordinates of the upper left corner, h and w are the height and width of R, respectively, in pixels;

[0035] Step 2: Calculate the set of rectangular bounding boxes RE={RE i |i=1,2,3,4}, so that it satisfies the following formula,

[0036]

[0037] Since the pixel width parameter ρ of the area around the car logo is too small, it is easily affected by noise, and if it is too large, the accuracy of judging the ...

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 sample expansion method fusing virtual samples, which is characterized in that the method comprises the following steps: firstly, the real sample of the collected vehicle logo is subjected to vehicle logo elimination processing to be made into a vehicle model library without a vehicle logo; secondly, noise and disturbance are added to the collected logo samples to enlargethe sample set, and then the logo is separated by a graph cutting method. In the third step, the separated vehicle logo and vehicle image are combined and fused to generate virtual samples. The invention has the beneficial effects that under the condition of small sample size, a new sample expansion scheme is proposed, and the method can obtain the diversity of vehicle objects to a greater extent, thereby improving the accuracy and adaptability of identification, especially for some uncommon samples with less existing sample libraries.

Description

technical field [0001] The invention relates to a sample expansion method for fusing virtual samples by using computer vision technology, in particular to a method for increasing the number of vehicle logo samples. Background technique [0002] With the development of the economy, the number of cars has increased dramatically, and this phenomenon has brought many social problems, especially the traffic problems in the first-tier developed cities. These problems are concentrated in road congestion, frequent traffic accidents, and repeated violations of vehicle regulations. In the process of dealing with traffic problems, the intelligent identification of illegal vehicles is particularly critical. As an important symbol of vehicles, vehicle logos play an important auxiliary role in vehicle identification. At present, deep learning technology is mostly used at home and abroad to identify car logos, and in this technology, the collection and expansion of car logo samples has be...

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): G06T11/60G06T3/40
CPCG06T3/4038G06T11/60
Inventor 高飞刘浩然林俊辉卢书芳张元鸣陆佳炜张永良肖刚
Owner ZHEJIANG 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