Object detection: Automatic license plate detection using deep learning and openCV

Abstract

Object Detection is one of the most important concepts of Computer Vision which is used in various areas like Medical Field, Security, Self Driving cars, Automated vehicle systems etc.We choose the application of Automatic License plate Recognition. Automatic License Plate Recognition is an emerging technology which is helpful in many fields and at the same time is challenging. It’s challenging because we need to get the accurate recognition of the characters in a number plate. In practical applications where sometimes the images are captured in the worst weather condition, bad lighting, wind. And to the addition, license plates are often dirty or blackened due to the smoke, half broken, or having scratches on certain characters and detection of too many license plates in a single frame. All these will act as the obstacles in developing an effective ALPR system. So basically, this is a system where recognition of characters from images using Computer Vision techniques are performed. This system is implemented in many fields like parking lots, private and public entrances, toll gates, theft control, checking the authentication of a vehicle. Procedure followed in this paper are, first capturing images from camera then loading that into system, preprocessing done using OpenCV library. Then we use Attention OCR a deep learning model to recognize the characters from an image. And later display that in the GUI and store them in the databases for different operations later.