Submitted to: ASABE Annual International Meeting
Publication Type: Abstract Only
Publication Acceptance Date: 6/20/2010
Publication Date: 6/20/2010
Citation: Kang, S., Kim, M.S., Lee, K., Son, J. 2010. Hyperspectral imaging for nondestructive evaluation of tomatoes.[abstract] . ASABE Annual International Meeting. Interpretive Summary:
Technical Abstract: Machine vision methods for quality and defect evaluation of tomatoes have been studied for online sorting and robotic harvesting applications. We investigated the use of a hyperspectral imaging system for quality evaluation and defect detection for tomatoes. Hyperspectral reflectance images were acquired for normal tomatoes spanning a range of ripeness and for defective tomatoes using a portable hyperspectral reflectance imaging system. Optimal wavelengths were determined by correlation analysis on intensity ratios and differences calculated for pairs of wavelengths. In order to distinguish tomatoes of three ripeness grades and also defective tomatoes, Principal Component Analysis (PCA) and correlation analysis for the wavelengths were used to analyze the ripeness and defect features. This research demonstrated that the hyperspectral imaging technique is useful for nondestructive determination of tomato quality.