License plate Recognition for moving vehicles case : At night and under rain condition

Research for the detection and identification of existing license plate focus on static image under different weather conditions. This research aims to detect and recognize the license plate on moving vehicle at night and under rain condition. The method used to detect license plate is Cascade Classifier with Local Binary Pattern (LBP) as a descriptor, top-hat transform for image enhancement, and recognizing character of license plate detected is using Optical Character Recognition (OCR) and Template Matching (TM). The testing process using 9 videos data with duration 30 seconds. The testing result that accuracy rates LBP OCR reach 77.42% increase 32.26% than without top-hat transform and LBP TM 64.52% increase 45.16% than without top-hat transform.