Convolutional Neural Network Based Intelligent Handwritten Document Recognition

This paper presents a handwritten document recognition system based on the convolutional neural network technique. In today’s world, is rapidly attaining attention of researchers due to its promising behavior as assisting technology for visually impaired users. also helpful automatic data entry system. proposed prepared dataset English language character images. The has been trained large set sample and tested images user-defined documents. this research, multiple experiments get very worthy results. will first perform image pre-processing stages prepare training using network. After processing, input segmented line, word segmentation. accuracy during segmentation up 86%. Then these characters are sent their recognition. technique in providing most acceptable accurate results given dataset. work approaches result 93%, validation that slightly decreases with 90.42%.