Performance evaluation of digital filters for Yorùbá optical character recognition systems
1 Department of Electrical, Electronic Engineering, The Federal Polytechnic, Offa, Nigeria.
2 Department of Electrical, Electronic Engineering, University of Ilorin, Nigeria.
3 Department of Computer Engineering, University of Ilorin, Nigeria.
Research Article
GSC Advanced Engineering and Technology, 2022, 03(02), 018–025.
Article DOI: 10.30574/gscaet.2022.3.2.0035
Publication history:
Received on 01 April 2022; revised on 15 May 2022; accepted on 18 May 2022
Abstract:
Computer vision systems largely depend on the quality of output of the image processing modules to perform their operations for a desired accurate result. The process of image acquisition and transmission usually results in image degradation. This endangers the efficiency of computer vision systems. This paper presents the causes of image degradation and the restoration techniques to enhance the output of computer vision systems. At different filter kernel sizes, median filters have better performance in image restoration as shown in the SNR, PSNR, and MSE results obtained. Averaging filters result in a blurring effect on the image. Wiener filters perform better for speckle and Gaussian noise. For impulse (salt and pepper) noise, median filters have the best performance.
Keywords:
Convolution; Degradation; Filter kernel; Mean squared error (MSE); Peak signal to noise ratio (PSNR); Restoration
Full text article in PDF:
Copyright information:
Copyright © 2022 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0
