Handwritten Character Recognition using Neural Network for Encryption System

Traditionally, handwritten character recognition has been done offline. More specifically, the writing was recorded optically by a scanner and used as an input image. However, there were issues with accuracy, and it was time-consuming. Online handwritten character recognition was established to combat this, which uses neural networks and is significantly superior to the previous method. Moreover, the main focus of this research paper is on cybersecurity. According to RiskBased Security analysis, data breaches exposed 36 billion records in the first three quarters of 2020. Despite this, the number of publicly publicized breaches has reduced by 51% compared to the previous year. According to research by Deep Instinct, malware usage jumped by 358 % through 2020, while ransomware usage increased by 435 percent from the prior year. Compared to the same month in 2019, malicious activity increased by 653 percent in July 2020. Thus we need to cope with better and more efficient cyber security methods, i.e., new encryption techniques.