CSIT - 2021 рік

Постійне посилання зібрання

Зараз показуємо 1 - 1 з 1

Neural network based image recognition method for smart parking
(Khmelnytskyi National University, 2021) Pavlova, O.; Kovalenko, V.; Hovorushchenko, T.; Avsiyevych, V.; Павлова, О.; Коваленко, В.; Говорущенко, Т.; Авсієвич, В.
Currently, the issue of creating smart parking lots is extremely important due to the rapid growth of number of cars, especially in big cities. Thus the need for parking spaces and search facilities still remains an urgent problem. Assuming that every day the average motorist spends 20 minutes searching for such a place, this is about 120 hours a year, which could be spent on something more useful. Today, there are many projects of "smart" parking, but practical examples can be counted on the fingers, and information about the cost-effective aspect of their implementation is generally very limited. The paper provides analysis of the most common methods and tools for smart parking and proves the advantages of camera-based method. The research in general is aimed at image recognition for camera-based smart parking using convolutional neural network.