Vehicle License Plate Localization and License Number Recognition Using Unit-Linking Pulse Coupled Neural Network

摘要

In this paper, Unit-linking Pulse Coupled Neural Network (U-PCNN) is applied in vehicle license Plate Localization (PL) and license Number Recognition (NR) being part of optical character recognition(OCR). PL and NR are cores of License Plate Recognition (LPR) system. In PL, firstly, the proposed algorithm based on U-PCNN edge detection highlights plate regions, and then using those results obtains plate locations. In NR, employing U-PCNN extracts features of license number. The experimental results show that the license plates properly extracted were 224 over 233 input images (96.137%) and the NR accuracy is 96.67%.