REVIEW PAPER Wavelet Analysis of Signals in Agriculture and Food Quality Inspection Chandra B. Singh & Ruplal Choudhary & Digvir S. Jayas & Jitendra Paliwal Received: 30 March 2008 / Accepted: 15 May 2008 / Published online: 17 June 2008 # Springer Science + Business Media, LLC 2008 Abstract Food quality and safety have become the top priorities for agriculture and food processing industry due to the increasing consumer demand for high-quality healthy food. The food processing industry is currently focusing on using fast, precise, and nondestructive automated quality inspection techniques. Near-infrared spectroscopy, image processing, hyperspectral imaging, X-rays, and ultrasonic techniques have been researched and shown to have high potential for automated inspection. The biggest challenge in the automated inspection systems deals with signal pre- processing, denoising, feature extraction, and its re-synthesis for classification purposes. Several research studies have established that the technique of wavelet analysis can very well resolve these issues of signal processing in many systems used for quality inspection of agricultural and food products. The objective of this paper is to discuss the theory of wavelet analysis and review its application in signal processing and feature extraction for quality monitoring of agricultural and food products. Keywords Wavelet analysis . Food quality . NIR spectroscopy . Image processing . Hyperspectral imaging . X-ray imaging Introduction Wavelet analysis of signals is increasingly becoming a popular tool in signal processing. Signal processing includes noise removal, compression, feature extraction, and reconstruction. The processing and analysis of signals of interest vary according to the discipline and objectives of a study. In agricultural and food products handling and processing industry, food quality and safety are the most important aspects to be addressed. Many of the advanced agricultural and food products quality inspection techniques use ultrasound signals, dielectric signals, nuclear magnetic resonance spectra, optical spectra including ultraviolet, visible and near infrared (NIR) bands, and digital image signals. Digital image signals can be two dimensional signals such as optical visible images, ultrasonic images, X- ray images or three dimensional signals such as hyper- spectral images, computed tomography (CT) images, and magnetic resonance images (MRI). Most of these signals need to be analyzed to extract useful information incorpo- rated within them. These signals can be either in analog or digital format, but in recent years digital signal processing has become popular. With the widespread availability of inexpensive digital processors and computers, most signals are obtained in or converted to digital format. However, due to the transient nature of signals produced by biological materials, traditional methods such as Fourier transform or short-time Fourier transform are unable to process such signals effectively (Marchant 2003). Wavelet transform has been recognized as a valuable tool in chemometrics and signal processing (Jetter et al. 2000). Compressing a large data set by wavelet transformation before regression is faster compared to directly applying the partial least square (PLS) regression on the original data (Trygg and Wold 1998). Apart from compression, wavelet transform can be successfully applied for separation of overlapping bands, noise removal, smoothing, base line correction, and in removing multicollinearity effect of multidimensional spec- tra (Barclay et al. 1997; Gributs and Burns 2006). Wavelet transformation of digital images discriminates several spatial orientations and it is very effective in analyzing the information content of images, texture discrimination, Food Bioprocess Technol (2010) 3:2–12 DOI 10.1007/s11947-008-0093-7 C. B. Singh : R. Choudhary : D. S. Jayas (*) : J. Paliwal Department of Biosystems Engineering, University of Manitoba, Winnipeg, MB R3T 5V6, Canada e-mail: Digvir_Jayas@umanitoba.ca