Vehicle Detection, Classification and Counting

• Published in: 2023 International Conference on Computational Intelligence, Communication Technology and Networking (CICTN) pp. 631 - 636
• Main Authors: Mali, Pashupathi, Pallavi, L., Kondakalla, Bhoomika, Ch, Madhu Babu, Y C A, Padmanabha Reddy, Kondaveeti, Bhargav
• Format: Conference Proceeding
• Language: English
• Published: IEEE 20.04.2023
• Subjects: and counting; and grey scale; canny algorithm; classification; streaming media; traffic surveillance; Vehicle detection; video processing
• DOI: 10.1109/CICTN57981.2023.10140443

According to logic, turnpike chiefs are becoming increasingly dependent on the differentiating evidence and counting of watchful automobiles. However, due to the wide range of vehicle sizes, discovery is still difficult, which affects the accuracy of the vehicle count. The review includes a further excellent thruway vehicle dataset with around 6000 clarified events in 12659 images. In contrast to existing publicly available datasets, the suggested dataset retains clarified tiny picture-related components, providing comprehensive information for deep learning-based vehicle detection. The proposed vehicle location and counting framework eliminates the parkway street surface from the image and divides it into a remote and proximal zone using a previously proposed division strategy; the method is essential for advancing vehicle discovery. Finally, video handling is used to obtain driving directions. A few parkway observation recordings from various scenes are used to validate the suggested procedures. Results from tests show how applying the suggested division technique might improve identification accuracy, especially for small vehicle items. Additionally, the original methodology described in this paper does a commendable job of determining driving route and counting automobiles. This paper has broad implications for the board and control of the Parkway Scene.