EXTENDED FREEMAN CHAIN CODE TECHNIQUE FOR LICENSE-PLATE RECOGNITION(LPR) USING MORPHOLOGICAL BASED METHOD(MBM)

This paper presents a MBM for extracting LPR from images. The MBM is proposed to extract features from LP. The features are tranformations of image scaling, translation and skewing. The LP is pre-processed, fragmented into several parts, and then, a LPR algorithm is applied. The LP verification is to check the number of characters in the plate. The MBM reduces the number of candidates in the LP. Nearly 130 examples were tested out of which 128 are recognized successfully. The average accuracy of LPR is 98%. Experimental results show that the proposed method performs very effectively for LPR. There are two tollgates situated at 10 and 20 KMs of distance. When the vehicle is passing the first tollgate, each vehicle is verified by sending the registration number to the particular local area police station the informations about the vehicle such as vehicle owner, his photograph, sex, address proof, number of cases on the vehicle, how many peoples are permitted to drive the vehicle alternatively, and their particulars, make of year, engine capacity, engine number, color of the vehicle, seating arrangements, vehicle dimensions, permit typeown use , State Permit, National Permit, Educational, Task details, periodic fitness certificate of the vehicle, RTO address etc., And the system will get reply before the vehicle is crossing the second tollgate. For doing all the transmission very fast, an Extended Freeman chain code is introduced and RLC technique is used to provide effective communication. In this paper, an extended 8x8 and 16x16 zoning feature is introduced to retain the local characteristics of a character.