Skip to main content
ARS Home » Midwest Area » East Lansing, Michigan » Sugarbeet and Bean Research » Research » Publications at this Location » Publication #212658

Title: On-Line Hyperspectral Transmittance Imaging for Internal Defect Detection of Pickling Cucumbers

Author
item ARIANA, DIWAN - USDA/FAS
item Lu, Renfu

Submitted to: ASABE Annual International Meeting
Publication Type: Proceedings
Publication Acceptance Date: 5/15/2007
Publication Date: 6/17/2007
Citation: Ariana, D., Lu, R. 2007. On-Line Hyperspectral Transmittance Imaging for Internal Defect Detection of Pickling Cucumbers. ASABE Annual International Meeting. Paper No. 073133.

Interpretive Summary: Internal damage to pickling cucumbers often takes place as a result of mechanical stress exerted on them, in the form of static or dynamic load, during harvesting and postharvest handling operations. Moreover, cucumbers that grow too fast or are harvested too late tend to have soft tissue with large seed cavity. Cucumbers with internal defect are susceptible to bloating during brining, which could result in soft and/or hollow pickle products that will be rejected by consumers. Hence it is desirable and necessary to separate these defective cucumbers at the early stage of sorting and grading. The current pickling cucumber grading and sorting method is largely based on size and shape, and it may have limited capabilities of detecting surface defects such as cuts or broken pieces and misshapen fruit. A machine vision-based system that is capable of identifying internal characteristics of cucumbers is thus needed for detecting and removing cucumbers with internal defect. We developed a laboratory prototype using hyperspectral transmittance imaging technique. Hyperspectral imaging provides both spectral and spatial information about individual cucumbers, and hence it can be more effective in detecting minor or subtle differences in the fruit tissue than conventional imaging technique which does not offer spectral information. The prototype was tested on fresh pickling cucumbers, either with or without internal defect, at a speed of approximately two fruit per second. Computer algorithms were developed to differentiate normal cucumbers from defective cucumbers resulting from mechanical stress due to rolling load that would be encountered during harvest and transport. The prototype achieved good classification between defective cucumbers and normal cucumbers, with an overall accuracy of 91.5%. With improvements in imaging and computation speed, the hyperspectral transmittance imaging technique could be implemented for online grading and sorting of fresh pickling cucumbers for internal quality/defect. The technology can also be useful for inspecting internal quality and/or defect of other fruits and vegetables to assure consistent and superior final products delivered to the consumer.

Technical Abstract: Hyperspectral imaging technique under transmittance mode was investigated for detection of internal defect in pickling cucumbers such as carpel suture separation or hollow cucumbers caused by mechanical stress. A prototype of on-line hyperspectral transmittance imaging system was developed for real-time detection of internal defect in pickling cucumbers. Experiments were conducted on ‘Journey’ pickling cucumbers, some of which were subjected to mechanical stress to induce internal defect in seed cavity. Partial least squares discriminant analysis (PLS-DA) and the use of single or ratio of two wavelengths were performed on mean and standard deviation spectra extracted from the hyperspectral images to predict the presence of internal defect. Internal defect on pickling cucumbers was best predicted in the region of 700-1000 nm with a classification accuracy of 91.5% using PLS-DA. The study shows that the proposed hyperspectral imaging technique can be used for internal defect detection.