Paper Title
Number Plate Character Extraction Using CNN

Abstract
Using license plate recognition technology, the vehicle registration number can be automatically recognized and retrieved, helping to gather information about the vehicle in an emergency. However, license plate recognition can be very difficult, especially in foggy and foggy conditions, where fog and mist can severely blur the lines and lettering on the license plate, making it difficult to identify. Various single-image dark channel de-fogging algorithms aim to solve the problem of image blur in a fast and efficient manner. Such algorithms rely on prior dark channel theory to estimate airglow, which serves as a key parameter for haze removal. This paper describes the use of Sobel technique and image formation for car license plate recognition and OCR technique is used to extract car number. Keywords - Vehicle number plate, Vehicle License, dehazing, image morphing, ocr