A vehicle detection method based on deep learning

A technology of deep learning and vehicle detection, which is applied in the field of vehicle detection based on deep learning, can solve problems such as image-dependent lighting and quality, and achieve the effects of reducing image preprocessing, high detection accuracy, and simple algorithms

Active Publication Date: 2021-09-17
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] Purpose of the invention: In order to solve the problem that the existing vehicle detection algorithm relies too much on the illumination and quality of the image, so that the vehicle detection has better adaptability and applicability, the present invention provides a vehicle detection method based on deep learning

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 vehicle detection method based on deep learning
  • A vehicle detection method based on deep learning
  • A vehicle detection method based on deep learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0037] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0038] The invention belongs to the technical field of deep learning, and in particular relates to a vehicle detection method based on deep learning, which uses the feature map of the last convolutional layer in the deep network to add weights to realize precise positioning of the vehicle. Method flow see figure 1 with figure 2 .

[0039] Step S1: The vehicle database with vehicle brand label information is de-meaned, and the deep learning model trained in the ImageNet database is selected as the basic network model. On the basis of the basic network model, the vehicle database with de-averaged value is used for fine-tuning training Network to obtain the final trained deep learning network for vehicle detection. Deep learning network model see image 3 .

[0040] The fine-tuning parameters are set as follows: the learning rate of the fixe...

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 vehicle detection method based on deep learning, in particular, a method of adding weights of feature maps of the last convolutional layer in a deep network to realize precise vehicle positioning. It belongs to the field of computer vision technology. The present invention first uses the vehicle database to train the deep learning network, then sends the picture to be detected into the trained network, obtains the category information of the picture through a forward propagation, and obtains the weight with the largest weight among the parameters according to the category information, and the last one The feature map of the convolutional layer is superimposed, and then image fusion is performed with the picture to be detected, and finally the accurate positioning of the vehicle is realized. This algorithm has good accuracy and adaptability. It effectively solves the problems of environmental interference, light influence, obstacle influence and low accuracy when traditional image processing algorithms realize vehicle detection, and can be applied to vehicle detection in different scenarios.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a vehicle detection method based on deep learning, which uses the feature map of the last convolutional layer in the deep network to add weights to realize precise positioning of the vehicle. Background technique [0002] In recent years, with the continuous growth of the economy, the number of cars has also continued to increase, which has caused a series of traffic problems. In order to solve these problems, intelligent transportation systems have become a research hotspot. The vehicle detection is the most important link in the intelligent transportation system. Accurate positioning of the vehicle position plays a key role in the research fields of vehicle counting and vehicle classification in the intelligent transportation system. [0003] Currently, vehicle detection algorithms mainly include feature-based vehicle detection algorithms, vision-based vehic...

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 Patents(China)
IPC IPC(8): G06K9/00G06K9/46G06K9/62
CPCG06V20/584G06V10/44G06V2201/08G06F18/2415
Inventor 孙涵阮航
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
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