期刊名称：International Journal of Advances in Engineering and Management
电子版ISSN：2395-5252
出版年度：2021
卷号：3
期号：6
页码：654-660
DOI：10.35629/5252-0306450459
语种：English
出版社：IJAEM JOURNAL
摘要：In recent years, with the emergence of Artificial Neural Networks (ANN), deep learning has become increasingly popular in a variety of disciplines, including surveillance, health, medicine, sports, robotics, and drones. Convolutional Neural Network (CNN) is at the forefront of outstanding developments in deep learning, combining Artificial Neural Network (ANN) and previous deep learning methodologies. Handwritten digits recognition in a computer vision system can be a difficult operation, but it's crucial to a variety of new applications.Machine learning and computer vision researchers used to use it substantially towards development and maintaining practical applications including certain automated check number reading. The aim of this study is to examine the performance of CNN with different number of hidden nodes and periods in the classification of handwritten digits. The CMN's Modified National Institute of Standard and Technology (MNIST) dataset is specified by the framework for this performance assessment. The MNIST database shall collect 60,000 digit training pictures and 10,000 digit analysis images. Our main objective is to develop an image recognition humanmade neural network.
