Determination of license plate location with deep learning

In this study we implemented license plate recognition using CNN. The study applied in two main steps. In the first step we detected license plate position with R-CNN. In the second step license plate reading operation was realized from detected license plate. License plate reading operation was performed with image processing. We used 450 license plate images for training R-CNN. For testing R-CNN model we used 100 license plates to images. In the second step we performed license plate reading from detected license plate. During the reading process noise of plate images removed and the image separated into characters.