Design of a field programmable gate array-based platform for real-time de-noising of optical imaging signals using wavelet transforms J. Carletta a, * , G. Giakos a , N. Patnekar b , L. Fraiwan b , F. Krach a a Department of Electrical and Computer Engineering, Auburn Science and Engineering Center, E. Buchtel Mall, The University of Akron, Akron, OH 44325, United States b Department of Biomedical Engineering, Olson Research Center, The University of Akron, Akron, OH 44325, United States Available online 7 October 2004 Abstract In this study, the design of a low-cost, field programmable gate array (FPGA)-based digital hardware platform that implements wavelet transform algorithms for real-time signal de-noising of optical imaging signals is presented. The proposed digital hardware de-noising platform can achieve a throughput of 25 Msamples per second, with a latency from input to output of 1.5 ls. The system fits on a 10K20, one of the smaller field programmable gate arrays from the Altera Flex 10K family, and represents a low-cost approach to signal de-noising. Preliminary experimental results clearly indicate the potential of the presented FPGA-based digital platform for real-time de-noising of signals utilized for the detection and imaging of surface defects and discontinuities. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Wavelet transform; Signal de-noising; Real-time computation 1. Introduction Several researchers have explored the use of wavelets for signal de-noising. Wavelets have been used to reduce noise in biomedical signals [1], in medical images [6], in digital communications channels [5], and in sonar signals [4], among oth- ers. Signal de-noising using wavelet transforms consists of the following three steps [9]: 1. Taking the wavelet transform of the noisy signal. 2. Thresholding the detail coefficients of the trans- formed signal. 0263-2241/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.measurement.2004.09.007 * Corresponding author. Tel.: +1 330 972 5993; fax: +1 330 972 6487. E-mail address: carlett@uakron.edu (J. Carletta). Measurement 36 (2004) 289–296 www.elsevier.com/locate/measurement