Handwritten Digit and Character Recognition UsingNeural Networks

  • Sheba Selvam, Megaldon Jasper K, Prashanth N, Preetham M, Manoj V

Abstract

IndividualPopularproblemincomputervisionandMachineLearningapplication isrecognitionofhandwrittendigitsandcharactersofEnglishhandwrittentextand different people will have different handwriting so it is difficult to identify them. Since the use of computer has a massive growth, the need to implement the handwrittencharactersordigitsofhumanbeingstoberecognizedbythecomputer or system. To solve this handwritten digit and character recognition problem, many Machine Learning techniques are used. Since CNN yields better output accuracy compared to other machine learning techniques, in this research we use neural network approaches of CONVOLUTIONNN. The proposed work mainly focusses on performance and rate of a recognition accuracy. Though only these are not main criteria in process of evolution, but also focus on executiontime.Toconductthisexperiment,itusesastandarddatasetofarandom handwrittendigits. In this research analysis we try to identify and recognize the handwritten digit patterns of either digit or characters of English handwritten text by using python and tensor flow by using Convolution Neural Network.

 

Published
2020-06-01
How to Cite
Sheba Selvam, Megaldon Jasper K, Prashanth N, Preetham M, Manoj V. (2020). Handwritten Digit and Character Recognition UsingNeural Networks. International Journal of Advanced Science and Technology, 29(06), 7497 - 7507. Retrieved from https://sersc.org/journals/index.php/IJAST/article/view/23959