POLICE NUMBER IDENTIFICATION USING ARTIFICIAL NEURAL NETWORK (LEARNING VECTOR QUANTIZATION)

Keyword: Keywords: image, LVQ, Pattern Recognition, Police Number Abstrack: ABSTRACT:Pattern recognition technology is widely used in various applications. One area of applying pattern recognition is biometrics. Pattern recognition system is a technology used to identify a person's identity through body parts or behaviors. In this paper the authors try to apply pattern recognition to identify the number of police cars. The introduction of the vehicle identification number of the police is very useful especially in the parking system. The system is still a lot of parking now that the manual way of record police numbers every car made by officers. This study aims to simulate the image processing methods to identify the number of police cars to replace the manual system made by humans. Input photograph image generated by digital cameras. The image is processed through a pre-processing for image has better quality and better prepared for the processing of feature extraction. Pattern recognition process is then performed on the image to get a feature found on that image. Features obtained from image processing in the form of vectors that represent the characteristic features of a letter or number. Characteristic features are then tested for the classification using Artificial Neural Network Learning methods Vector quantization (LVQ). LVQ to classify the test vectors by comparing with the weight of each class so as to produce a distance. The distance will determine the class of the tested traits. This research was only done on the car plate number. Experiments conducted by collecting data in the form of photos police numbers. From the data collected and then grouped by character letter or number to make a template. The results showed the success of 78% of known character patterns and the unknown.