An implementable system for detection and recognition of license plates in Pakistan

the focal idea of following paper is to introduce a system which is more usable and efficient to detect and recognize the vehicle number plate in Pakistani scenario. There are always some constraints in which an automatic number plate recognition (ANPR) system work. The challenges include poor illumination, camera speed and variable background. In this paper, we propose a method where distinct feature matching is applied to get acceptable results in Pakistani conditions where number plates usually don't follow a standard pattern. The developed system is tested on a dataset of fifty images containing variable size and style number plates having Latin script. The initial results show that our system is able to detect and recognize most of the plates and can be tested further to see the feasibility of implementing it in real environment with multi-script recognition.