Skip to main content
ARS Home » Midwest Area » East Lansing, Michigan » Sugarbeet and Bean Research » Research » Publications at this Location » Publication #329676
Research Project: Nondestructive Quality Assessment and Grading of Fruits and Vegetables

Location: Sugarbeet and Bean Research

Title: Structured-illumination reflectance imaging as a new modality for food quality detection

Author
item LU, YUZHEN - Michigan State University
item Lu, Renfu
item LI, RICHARD - Michigan State University

Submitted to: Meeting Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 6/2/2017
Publication Date: 7/1/2017
Citation: Lu, Y., Lu, R., Li, R. 2017. Structured-illumination reflectance imaging as a new modality for food quality detection. In: CIGR-AgEng Conference Proceedings, June 26-29, 2017, Aarhus, Denmark. Paper No. 544.

Interpretive Summary:

Technical Abstract: Uniform or diffuse illumination is the standard in implementing many different imaging modalities. This form of illumination, however, has some major limitations in acquisition of useful information from food products because reflectance from the food products is non-uniform due to irregular, curved product surfaces. Moreover, diffuse illumination is not suitable for detecting the structural features of tissues. This paper reports on a new structured-illumination reflectance imaging (SIRI) system for enhanced detection of defects in food products. The technique utilizes special patterns of illumination for controlling light penetration in food, and then uses image demodulation to create depth-resolved images. The system mainly consisted of a high-performance digital camera and a light pattern generating projector. It was implemented with sinusoidal patterns of illumination with different spatial frequencies for detecting foreign objects embedded in a diffusive nylon block and fresh bruises in apples. Results showed that depth-resolved features detection depended on the spatial frequency of sinusoidal illumination; higher frequency resulted in less penetration of light in the tissue but a higher spatial resolution for better identification of foreign objects or bruises in the samples. Three phase-shifted images were needed for effective demodulation of direct component (DC) images, representing regular images under uniform illumination, from amplitude component (AC) images, representing the effect of the alternating component. Fresh bruises in apples were effectively identified using the SIRI technique, which, otherwise, could not be achieved using uniform illumination. The research demonstrated that SIRI can enhance the detection of defects in fruit, and it is promising for food quality evaluation.