Using composite sinusoidal patterns in structured-illumination reflectance imaging (SIRI) for enhanced detection of defects in food

Authors: LU, YUZHEN - Michigan State University; Lu, Renfu

Submitted to: Journal of Food Engineering
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/16/2016
Publication Date: 12/18/2016
Citation: Lu, Y., Lu, R. 2016. Using composite sinusoidal patterns in structured-illumination reflectance imaging (SIRI) for enhanced detection of defects in food. Journal of Food Engineering. 199:54-64.

Interpretive Summary: A structured-illumination reflectance imaging (SIRI) system was recently developed in our laboratory for food quality detection. The technique was demonstrated to be superior to popular uniform-illumination imaging technique for detecting subsurface defects in food like bruise. However, conventional implementation of the technique requires applying sinusoidal patterns of illumination for single spatial frequencies, which could limit its ability of interrogating the tissues at different depths because light penetration is dependent on spatial frequency. This study, for the first time, explored the feasibility of using composite sinusoidal patterns that integrate two or three spatial frequencies into one illumination pattern, to allow better detection of tissues at different depths. Three mathematical methods based on Fourier transform were proposed to retrieve amplitude component images for the acquired images, which is critical for obtaining depth-resolved information for the samples. These methods were phase-shifting with or without spiral phase transform and frequency-domain filtering. Numerical simulations were conducted to evaluate the three methods for dual-frequency and triple-frequency composite sinusoidal patterns, followed with an experiment on detection of fresh bruises in ‘Gala’ and ‘Golden Delicious’ apples using the SIRI system. Results from simulation and experiment showed that the phase-shifting methods resulted in good performance as measured by image demodulation error and image contrast for bruised tissue detection. The filtering-based method, although viable in numerical simulation, had worst performance among the three methods. Composite sinusoidal patterns of illumination have some advantages, compared to single sinusoidal patterns, in detecting defects that may occur in food at varying depths.

Technical Abstract: Structured-illumination reflectance imaging (SIRI) provides a new means for enhanced detection of defects in horticultural products. Implementing the technique relies on retrieving amplitude images by illuminating the object with sinusoidal patterns of single spatial frequencies, which, however, are limited in interrogating the tissues at different depths. This study presented a first exploration of using composite sinusoidal patterns that integrated two and three spatial frequencies of interest, with SIRI for enhanced detection of subsurface defects in food (e.g., bruises in apples). Three methods based on Fourier transform were proposed to retrieve amplitude images at different frequencies by using either phase shifting with or without spiral phase transform (SPT) or frequency-domain filtering. The phase-shifting method involves solving a linear system that is composed of multiple phase-shifted pattern images in the Fourier space, and SPT, which acts as a two-dimensional quadrature transform operator, is used to reduce the images needed for amplitude retrieval; while the filtering method directly extracts different frequency components from only one pattern image, which are then subjected to SPT processing. The three methods were tested, using dual-frequency and triple-frequency composite sinusoidal patterns, in numerical simulations and experiments on the detection of fresh bruises in apples by SIRI. The phase-shifting methods showed good performance in terms of small demodulation errors and strong image contrast for bruise detection; the filtering-based method, although viable in numerical simulation, needed improvement due to the worst practical performance. In addition, more frequency components involved would deteriorate the performance of these methods, and grid composite patterns were superior to the fringe ones due to reduced interactions between different frequency components.