Combination of CNN-GRU Model to Recognize Characters of a License Plate number without Segmentation

Recognition is a genre of manipulation of digitized image automation for discovering the number plate details from a given image. Due to various factors, it is difficult to achieve great recognition results for the license plate. In general, human beings can easily read characters in license plate, but the machine cannot do until it is trained to do so. Now a day's vehicles are increasing day by day, to note down every vehicle plate number manually is difficult. To avoid that, optical character recognition (OCR) technology is used which extracts the license plate directly. In this paper, CNN (convolution neural network) -GRU (gated recurrent unit) model is developed. CNN is used for feature extraction and GRU is used for sequencing without using any segmentation methods. Finally, the character is recognized by utilizing a model design which is prepared on the dataset by GRU unit. A deep learning technique increases performance than traditional approaches like template matching. The testing precision of the proposed framework is 100% and training accuracy is 90%.