Skip to main content
ARS Home » Midwest Area » East Lansing, Michigan » Sugarbeet and Bean Research » Research » Publications at this Location » Publication #276302

Title: Development of automated inspection technology for quality grading of pickling cucumbers

item Lu, Renfu

Submitted to: Michigan State University Cucumber Reporting Session
Publication Type: Experiment Station
Publication Acceptance Date: 12/2/2011
Publication Date: 12/6/2011
Citation: Lu, R. 2011. Development of automated inspection technology for quality grading of pickling cucumbers. Michigan State University Cucumber Reporting Session. p. 44-50.

Interpretive Summary:

Technical Abstract: Pickling cucumbers are susceptible to external and internal damage during growth, harvest, transport, and postharvest handling. It is estimated that approximately 5-10% of harvested pickling cucumbers fall into the defect category. Pickling cucumber quality defect can occur in the form of soft or watery carpel tissue, loose seeds, split or hollow center, insect infestation, rot or scab, shrivel, misshapen fruit, translucent fruit, etc. Research has been conducted in the ARS East Lansing lab on developing hyperspectral imaging-based inspection technology for automatic sorting and grading of pickling cucumbers and pickles for external and internal quality defects. This report summarizes the research progress in measurement of the optical absorption and scattering properties of normal and defective pickling cucumbers, identification of optimal wavebands for internal defect detection, and detection of fruit fly infestation in pickling cucumbers using hyperspectral imaging technique. It was found that mechanical damage caused more significant changes in the optical scattering properties than in the absorption properties of pickling cucumbers. The optimal four wavebands identified in the visible and near-infrared region achieved the overall classification accuracy of 95% for pickling cucumbers and of 83% for whole pickles. Hyperspectral imaging in transmittance mode also achieved superior results (~90%) for detecting pest-infested cucumbers, compared with an overall accuracy of 75% by human inspectors.