Location: Sugarbeet and Bean ResearchTitle: Detection of subsurface bruising in fresh pickling cucumbers using structured-illumination reflectance imaging
|LU, YUZHEN - Mississippi State University|
|ZHANG, ZHAO - North Dakota State University|
Submitted to: Postharvest Biology and Technology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/11/2021
Publication Date: 6/17/2021
Citation: Lu, Y., Lu, R., Zhang, Z. 2021. Detection of subsurface bruising in fresh pickling cucumbers using structured-illumination reflectance imaging. Postharvest Biology and Technology. 180. Article 111624. https://doi.org/10.1016/j.postharvbio.2021.111624.
Interpretive Summary: Pickling cucumbers are susceptible to bruising during harvest and postharvest handling. Bruised or mechanically injured cucumbers should be removed prior to processing to ensure quality of final pickled products. Conventional imaging techniques using uniform illumination are widely used for food quality and safety inspection, but they are often ineffective for detecting subsurface defects like bruises. In this research, we used an inhouse developed new imaging system, called structured-illumination reflectance imaging (SIRI), for detection of subsurface bruising in pickling cucumbers. Instead of using uniform illumination with conventional imaging techniques, SIRI applies sinusoidal patterns of illumination at selected spatial frequencies to samples, to obtain pattern images. By using a mathematical method, called image demodulation, we can obtain two sets of images, i.e., direct component (DC), which are equivalent to images obtained under uniform illumination, and amplitude component (AC), which are unique to SIRI. In this study, pattern images at 700-900 nm were collected using SIRI for 240 ‘Vlaspick’ pickling cucumbers, half of which had been subjected to mechanical stress for inducing bruises. DC and AC images were then obtained through image demodulation for these cucumbers. AC images were found to provide more visually distinct features about bruises on cucumbers, which were either less distinct or invisible from DC images. Machine learning algorithms were developed, by using either all image features or selected features at 740 nm, for classifying the cucumbers into normal and bruised classes. Results showed that AC images and combination of DC, AC and their ratio images resulted in significantly better accuracies in detecting bruises, compared to DC images alone. By using the best five image features, SIRI was able to achieve as high as 96% overall classification accuracy. This research demonstrated the potential of SIRI for effective detection of subsurface defects like bruise in pickling cucumbers. With further improvements in the hardware and software design, the technique can be used for quality inspection of pickling cucumbers and other horticultural products.
Technical Abstract: Pickling cucumbers are susceptible to bruising during harvest and postharvest handling. It is thus desirable to segregate bruised fruit before they are marketed as fresh products or processed as pickles. Structured-illumination reflectance imaging (SIRI) is an emerging optical imaging modality for food quality and safety inspection. This study reported the first demonstration of SIRI for detecting subsurface bruising in fresh pickling cucumbers. Two independent sets of images, i.e., direct component (DC) and amplitude component (AC), were demodulated from the phase-shifted sinusoidal pattern images at 740 nm obtained for 240 pickling cucumbers. AC was found more effective than DC for ascertaining bruises that exhibited no visual symptoms. Classification models based on support vector machine were built using extracted image features, to classify cucumbers into bruised and normal classes. The highest classification accuracy of 91% was achieved by the ensemble of DC, AC and their ratio (AC/DC) images, which represented 7.6 percentage-point improvement over that obtained using the DC images alone. Using features selection for five sets of image features led to further improvements in the classification performance. Incremental evaluation of top 50 most informative features resulted in an averaged overall accuracy of 94%, with the highest accuracy of 97% attained by 31 features; and using a subset of only 5 features, 3 from AC and 2 from DC, also produced a high overall accuracy of 96%. This study showed that SIRI can provide an effective means for detection of subsurface bruising in pickling cucumbers. More research is, however, needed to implement SIRI for real-time inspection of cucumber defects.