Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Smart vehicle model identification method

A technology of model and car, applied in the fields of image search and urban smart transportation, it can solve the problems of inability to converge and increase the difficulty of training, and achieve the effect of reducing training cost and application difficulty.

Inactive Publication Date: 2018-09-04
JINAN INSPUR HIGH TECH TECH DEV CO LTD
View PDF8 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The larger the α value, the larger the expected object distance, but it may cause problems such as increased training difficulty and even failure to converge, so it is necessary to set the α value reasonably

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
  • Smart vehicle model identification method
  • Smart vehicle model identification method
  • Smart vehicle model identification method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0021] The content of the present invention is described in more detail below:

[0022] The present invention learns image features through a deep convolutional neural network, realizes the mapping from an image to a low-dimensional Euclidean space vector, and uses a special Triplet Loss function to train the deep convolutional network, so that the same type of car image in the Euclidean space The vectors have closer distances, enabling recognition of different car model categories.

[0023] The specific operation is as follows

[0024] 1. Collect image datasets of different car models. Mainly the frontal image of the car and the frontal image of the parking space. The car color can change, the model will ignore the car color, and confirm the car model through the structural design of the car head and tail, such as distinguishing between Audi Q3 and Audi Q5, and not distinguishing between black Audi Q5 and white Audi Q5. All images must have corresponding car model labels a...

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, which belongs to the technical field of urban smart traffic and image search, provides a smart vehicle model identification method. A vehicle image is mapped into a low-dimensional European space vector; and an Euclidean distance between vectors is calculated to determine a vehicle model. To be specific, a deep convolutional network is built; and the network is trained by using lotsof triple samples, so that the network has the capability of mapping of being lower than different types of vehicle picture distances by the same types of vehicle picture distances. The method has the characteristic of random application by training once; a problem of classification of vehicle models not appearing in training samples is solved; the training cost is lowered substantially; and theapplication difficulty is reduced.

Description

technical field [0001] The invention relates to urban intelligent traffic and image search technology, in particular to a method for identifying intelligent vehicle models. Background technique [0002] Automobile is an indispensable and important invention of human civilization. As the main means of transportation in today's society, it has played a huge role. However, with the development of the automobile industry, the appearance of automobile models is increasing, and the homogeneity is serious. Computer algorithms are used to identify automobile appearance pictures. Vehicle models become a difficult problem. [0003] With the development of artificial intelligence technologies such as deep learning, breakthroughs have been made in the classification and detection of image objects, and many innovative deep neural network models have been invented. The network models for image classification include convolutional neural network models such as LeNet, VGG, and ResNet, and ...

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): G08G1/015G06K9/62
CPCG08G1/015G06V2201/08G06F18/214
Inventor 高岩段成德于治楼
Owner JINAN INSPUR HIGH TECH TECH DEV CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap 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