A multi-attribute recognition method of vehicle's line-pressing in parking lot based on multi-task convolution neural network
by Shaohui Zhong; Ting Hu
International Journal of Information and Communication Technology (IJICT), Vol. 20, No. 3, 2022

Abstract: In order to solve the problems of low recognition accuracy and long recognition time, a multi-attribute recognition method based on multi-task convolution neural network is proposed. The structure principle of multi-task convolution neural network is analysed, and multi-task is set in the bottom area of convolutional neural network. The Hough transform is used to extract the parking line in the parking lot, and the input layer of the multi-attribute label structure is established by multi-attribute classification convolution neural network. The loss function of vehicle line pressing attributes in different parking lots is obtained by combining the full connection layer and the connecting sub layer. The multi-attribute recognition of vehicle pressure line is realised by measuring and learning the line voltage attributes of vehicles. The experimental results show that the method can effectively identify the line pressing situation of vehicles in parking lot, and the recognition accuracy can reach 99%.

Online publication date: Thu, 07-Apr-2022

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Information and Communication Technology (IJICT):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com