Handwriting Text Recognition using CNN and RNN

In the field of optical character recognition, there are still unans wered research questions regarding the recognition of handwritten text. In this paper, an effective method for developing handwritten handbook recognition systems is proposed This article uses a 3-subcaste Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) in conjunction with a supervised literacy technique. Although bit chart descriptions of the input samples boost the delicateness of any textbook recognition system, they are used as point vectors in the s ys tern. The objective vectors are pre-processed before the resulting goal variables according to input samples are applied to the CNN. Using samples of each digit in the number 123, the CNN&RNN training procedure is carried out to verify the system's general connection to new inputs. Two different algorithms for literacy are utilized in this study. Cumulative image processing techniques have also been developed to deal with the several characters that are provided in a single image, cocked image, and rotated image. The trained systern provides a better delicacy.