An Off-line Handwriting Recognition Employing Tensorflow
Hao Yu Zeng · 2020
Handwriting has been a conventional means of communication and recording in daily life since early time. Given its ubiquity in human transactions, machine recognition of handwriting has practical significance, such as, in reading handwritten notes in a PDA, in postal addresses on envelopes, in amounts in bank checks, or in handwritten fields in forms [1]. Handwriting recognition is a vital application in daily activities and the researches of especially handwritten digits recognition is vital. This paper focuses on using simpler neural network instead of complicated ones that require high quality of computer configuration to recognize handwriting digits with relatively promising accuracy. To do this research, a neural network to recognize handwriting in MNIST dataset using Softmax Regression algorithm with a high accuracy is built.