Skip to main content
ARS Home » Midwest Area » East Lansing, Michigan » Sugarbeet and Bean Research » Research » Publications at this Location » Publication #366165

Research Project: Nondestructive Quality Assessment and Grading of Fruits and Vegetables

Location: Sugarbeet and Bean Research

Title: siriTool: a Matlab graphical user interface for imaging analysis in structured-illumination reflectance imaging for fruit defect detection

Author
item LU, YUZHEN - US Department Of Agriculture (USDA)
item Lu, Renfu

Submitted to: ASABE Annual International Meeting
Publication Type: Proceedings
Publication Acceptance Date: 7/5/2019
Publication Date: 7/10/2019
Citation: Lu, Y., Lu, R. 2019. siriTool: a Matlab graphical user interface for imaging analysis in structured-illumination reflectance imaging for fruit defect detection. ASABE Annual International Meeting. 1900069. https://doi.org/10.13031/aim.201900169.
DOI: https://doi.org/10.13031/aim.201900169

Interpretive Summary: Structured-illumination reflectance imaging (SIRI) is an emerging modality for enhanced detection of defects in horticultural and food products. With this technique sinusoidal patterns of illumination at single or multiple spatial frequencies are applied to the object to obtain the so-called patterned images with two or three phase shifts. The acquired patterned images then go through an image demodulation process, a critical step in implementing the technique, to result in two independent sets of images, i.e., direct component (DC) and amplitude component (AC). DC images are equivalent to those acquired under diffuse, uniform illumination, while AC images are unique to SIRI, which contain special features that can reveal more useful information about the characteristics of subsurface tissues. In this study, a user-friendly graphical user interface, called siriTool, was developed, mainly based on the various methods developed by our group, to facilitate image processing and analysis. siriTool automatically performs all steps of image preprocessing (i.e., demodulation, background removal and image enhancement), features extraction and selection, and image classification. An application example was presented on using siriTool for the detection of yellowish subsurface spot defects in picking cucumbers. SIRI achieved more than 98% classification accuracies, which were significantly better than those obtained under conventional, uniform illumination. siriTool provides a user-friendly tool for automatic processing and analysis of SIRI images for defect detection of horticultural and food products.

Technical Abstract: Structured-illumination reflectance imaging (SIRI) is an emerging imaging modality that provides more useful discriminative features for enhanced detection of defects in fruit and other horticultural and food products. In this study, we developed a Matlab graphical user interface (GUI), i.e., siriTool, to facilitate image analysis in SIRI for fruit defect detection. The GUI enables image preprocessing (i.e., demodulation, background removal and image enhancement), features extraction and selection, and classification. The demodulation is done using a three-phase or two-phase based approach depending on image data acquired, whereas background removal is implemented based on automatic unimodal thresholding and image enhancement is achieved using fast bi-dimensional empirical decomposition followed by image reconstructions. For defect detection, features of different types are extracted from the enhanced images and features selection is then performed to reduce the feature set. Finally, the full or reduced set of features are input into different classifiers [e.g., support vector machine (SVM)] for image-level classifications. An example was presented on the detection of yellowish subsurface spot defects of pickling cucumbers. SIRI achieved over 98% classification accuracies based on SVM modeling with extracted features, which were significantly better than those obtained under uniform illumination.