Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Method and system for recognizing object type

A type recognition and object model technology, applied in the field of computer vision, can solve the problems of being easily affected by the environment, low recognition rate, low recognition performance, etc., to achieve the effect of improving the accuracy and improving the degree of discrimination.

Active Publication Date: 2015-06-03
SUZHOU KEDA TECH
View PDF5 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Active detection is to use the principle of reflected echo. The acoustic wave generator emits high-frequency waves, and then the moving vehicle returns at a changing frequency. The frequency characteristics are recorded by the transducer to identify the vehicle type. However, the shortcoming of this method lies in the implementation The process is easily affected by the environment, and the recognition performance decreases with the influence of ambient temperature and airflow
These two types of methods can only distinguish large, medium, and small vehicles, but cannot effectively distinguish them. For example, passenger cars, trucks, vans, cars, etc. are more detailed types of vehicles, and are easily affected by scenes, lighting, deformation, etc. The recognition rate is low
[0006] According to the above analysis, it can be seen that the existing object (especially for vehicles) type recognition methods have certain defects. Therefore, a technology for accurately and quickly identifying object types in complex scenes is needed.

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 and system for recognizing object type
  • Method and system for recognizing object type

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0028] figure 1 It shows an object type recognition method in Embodiment 1 of the present invention. This method can be applied to identify vehicle types in the scene of intelligent traffic checkpoints, for example, distinguishing whether a vehicle is a passenger car, a truck, a van or a car, etc., and can also be applied to In the field of home service robots, family members are identified, such as distinguishing whether a family member is an elderly person, a pregnant woman or a child. In short, it can be used in any field that needs to identify object types. The object type identification method comprises the following steps:

[0029] Step S11 , intercepting the area where the target to be identified is located from the acquired image. For example, extract the key frame image where the target to be identified is located from the video image, use Sobel (sobel) edge detection algorithm and vertical texture projection algorithm, and combine vehicle positioning technology to ...

Embodiment 2

[0054] In this embodiment, the object type recognition method of the present invention is described by taking the recognition of the vehicle type in the scenario of an intelligent traffic checkpoint as an example. The method may include the steps of:

[0055] Step A3, intercepting the area where the target to be identified is located from the acquired image. Specifically, the key frame image where the target to be recognized is extracted from the video image, and the Sobel (sobel) edge detection algorithm and vertical texture projection algorithm are used to locate the area where the target to be recognized is located. The proportion intercepts the area where the target to be recognized is located.

[0056] Step B3, enhancing the contrast of the image of the area where the target to be identified is located.

[0057] Step C3, performing scale normalization on the contrast-enhanced image of the area where the target to be identified is located.

[0058] Step D3, obtaining th...

Embodiment 3

[0066] In this embodiment, the identification method of the object type of the present invention is described by taking the identification of family members in the field of home service robots as an example. The method may include the steps of:

[0067] Step A4, intercepting the area where the target to be identified is located from the acquired image. Specifically, the key frame image where the target to be identified is extracted from the video image, and the Sobel (sobel) edge detection algorithm and vertical texture projection algorithm are used to locate the area where the target to be identified is located. The proportion intercepts the area where the target to be recognized is located.

[0068] Step B4, enhancing the contrast of the image of the area where the target to be identified is located.

[0069] Step C4, performing scale normalization on the contrast-enhanced image of the area where the target to be identified is located.

[0070] Step D4. Obtain the scale-n...

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 and a system for recognizing an object type. The method comprises the following steps: intercepting the area of a target to be recognized from an obtained image; obtaining the windowed Fourier transform characteristic pattern of the image in the area of the target to be recognized; utilizing a pre-training object model to detect the windowed Fourier transform characteristic pattern so as to obtain the object type. The windowed Fourier transform characteristic pattern is obtained by extracting windowed Fourier transform characteristics on the image, and then the windowed Fourier transform characteristic pattern is detected by utilizing the pre-training object model so as to obtain the object type, so that the discriminative degree of the object model for different objects can be improved, and the recognition accuracy rate for the object type is improved by 5%.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to an object type recognition method and system. Background technique [0002] In the field of intelligent transportation, vehicle recognition is a very important technology. Vehicle identification technologies currently on the market can be mainly divided into contact and non-contact. [0003] Contact vehicle identification is to bury the sensor in the ground, and the sensor will recognize the vehicle when it passes by. The disadvantage is that the construction is complicated, the installation process has a great impact on the reliability and life of the sensor, and the sensor is easily damaged by heavy vehicles, the maintenance cost is high, and the traffic needs to be interrupted during the installation and repair of the sensor, which will damage the road surface and affect the life of the road surface. In addition, in actual use, due to the influence of vehicle speed a...

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): G06K9/00G06K9/62G06N3/02
Inventor 赵鹏伦晋兆龙陈卫东
Owner SUZHOU KEDA 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
Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com