An Efficient Number Plate Detection System Based on Indian Traffic Rules

Abstract

With the invent of automobile and advances in the world of machinery, there is a notable increase in the number of vehicles, two wheelers and three wheelers in the road transport. With the raise in the number of vehicles there is a lot of traffic violations that is happening in the road system. Based on the Indian traffic road system, monitoring the violation of such a incident is a huge task and also a tedious system. But however steps are taken to monitor the traffic breach that happens in the road transport through License Plate Detection mechanism for the vehicles that involves in over speeding or violating the rules. License plate location is a very important concept in vehicle license plate recognition for intelligent transport systems. Number plates can have different shapes and sizes along with different colors. The most common vehicle number plate in India have the background color as yellow or white with the font color black. Identification of number plate for vehicles in India has been discussed in this paper and the numbers have been segmented to identify them specifically. Our focus here is on two main steps: first is to find the number plate and second the segmentation of the number to identify them specifically. The main objective of this paper is to efficiently design and implement a method for License Plate Recognition (LPR) of Indian License Plates. We have manually acquired the images of various vehicles. Here for the detection point, we have used COCO-API RCNN and K-NEAREST NEIGHBOR.