Cloud-based efficient scheme for handwritten digit recognition
- PDF / 1,081,642 Bytes
- 13 Pages / 439.37 x 666.142 pts Page_size
- 72 Downloads / 249 Views
Cloud-based efficient scheme for handwritten digit recognition Zeeshan Shaukat 1 & Saqib Ali 1 & Qurat ul Ain Farooq 2 & Chuangbai Xiao 1 & Sana Sahiba 1 & Allah Ditta 3 Received: 14 February 2020 / Revised: 3 July 2020 / Accepted: 29 July 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Handwritten character recognition has been acknowledged and achieved more prominent attention in pattern recognition research community due to enormous applications & vagueness in application methods, while cloud computing delivers appropriate, ondemand access of network to a joint tarn of configurable computing resource & digital devices. Principally two steps, feature extraction & character recognition, are required for Handwritten Digit Recognition (HDR), which are primarily based on some classification algorithms. Previous studies show the nonexistence of higher precision and truncated computational swiftness for HDR procedure. “The projected research aimed to make the trail towards digitalization clearer by providing high accuracy and faster cloud-based computational for handwritten digits recognition. The current study utilized a cloud-based neural network (CNN) as a classifier, suitable parameters of dataset MNIST for testing and training purposes as a framework called DL4J for cloud-based handwritten digit recognition. The said system magnificently managed to obtained precision up to 99.41%, which is higher than previously projected systems. Additionally, the proposed method decreases cost and computational time significantly as using cloud-based architecture for testing and training; as a result, the algorithm becomes more efficient. Keywords Convolutional Neural networks . Cloud computing . Handwritten digit recognition . Deep learning 4 J Highlights • Cloud-Based Handwritten Character Recognition System. • Cloud Based Novel Neural Network • 99.41% Accuracy Obtained, Decreases in Cost & Computational time
* Zeeshan Shaukat [email protected]
1
Faculty of Information Technology, Beijing University of Technology, Beijing 100124, China
2
Faculty of Life Science & Bioengineering, Beijing University of Technology, Beijing 100124, China
3
Division of Science & Technology, University of Education, Lahore 54000, Pakistan
Multimedia Tools and Applications
1 Introduction Cloud-Based Handwritten Digit Recognition is an open problem in computer vision and pattern recognition, and solving this problem has elicited increasing interest. In the realm of Optical Character Recognition (OCR), it is challenging to recognize characters from real or handwritten images or printed text image documents [12, 41]. The main challenge of this problem is the design of an efficient method that can recognize the handwritten digits that are submitted by the user via digital devices. The dealing of Handwritten Digit Recognition HDR is of huge apprehension in academic and commercial approaches [2, 4]. Scholars have been using various algorithms of machine learning to inspect handwritten digit recognition, w
Data Loading...
