Location: Healthy Processed Foods Research
Title: NIR detection of pits and pit fragments in fresh cherries (abstract)Author
![]() |
Liang, Peishih |
![]() |
Haff, Ronald |
Submitted to: Meeting Abstract
Publication Type: Abstract Only Publication Acceptance Date: 2/18/2016 Publication Date: 7/17/2016 Citation: Liang, P., Haff, R.P. 2016. NIR detection of pits and pit fragments in fresh cherries (abstract). ASABE International Meeting, July 17-20, 2016, Orlando, FL. lb. Interpretive Summary: Technical Abstract: The feasibility of using near infrared (NIR) diffuse reflectance spectroscopy for the detection of pits and pit fragments in cherries was demonstrated. For detection of whole pits, 300 cherries were obtained locally and pits were removed from half. NIR reflectance spectra were obtained in triplicate in the spectral range from 800 to 2100 nm using a standard commercially available spectrometer (Bruker, model MPA) and in the spectral range from 900 to 1700 nm using a handheld portable spectrometer (JDSU, model micro-NIR). For MPA data, cherries were correctly classified 99% of the time, with no false negative (fn) and one percent false positive (fp) results. For the JDSU data, cherries were correctly classified 91.6% of the time with 3% fp and 5.4% fn results. A second data set was tested which included spectra of cherries with no pits (460 spectra), whole pits (307 spectra), half pits (157 spectra), and quarter pits (138 spectra). Using MPA data, cherries with pits or pit fragments were distinguished from those without pits with 88% accuracy, with 13% fp and 11% fn. |