In the era of fast growing technologies, there is a huge demand among people for a comfortable lifestyle and travelling. In the last few years, the number of vehicles on road has been explosively grown. With the increasing growth in the vehicular sector every day, the need to track individual vehicle becomes a very challenging task. The proposed project suggests an automated way of tracking the fast moving vehicles with the help of the surveillance cameras on the road. One of the tedious step towards this project is to process the CCTV footage in the real time background. To overcome this strategy, an efficient deep learning framework particularly meant for object detection named YOLO You Only Look Once is used. The proposed work consists of four main steps. The first process is to convert the video footage into images and detect the car from each of the frames. The next step is to detect the license plate from the detected cars. Reading and recognizing the number plate characters from the detected number plates form the final step. To make the process of training the deep learning model easier, this system uses ImageAI library. It provides very efficient and easy to use classes to perform image recognition tasks. This system uses a customized dataset of 1000 images specifically containing Tamil Nadu license plate images. These images are taken under distinct light conditions and contradictory angles. The proposed method is tested in real time environment and achieved 97 accuracy for car detection, 98 accuracy for number plate localization and 90 accuracy for character recognition.
by P. Harini | K. Harish | V. Gnanaprakash "Real Time Number Plate Recognition"
Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-4 | Issue-5 , August 2020,
URL: https://www.ijtsrd.com/papers/ijtsrd31872.pdf
peerreviewedinternationaljournal, callforpaperinugcapprovedjournals, paperpublicationforstudent
