Click on “Download PDF” for the PDF version or on the title for the HTML version.
If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.
Ripeness Classification of Oil Palm Fresh Fruit Bunches Using Optical Spectrometer and Support Vector Machine
Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org
Citation: 2021 ASABE Annual International Virtual Meeting 2100287.(doi:10.13031/aim.202100287)
Authors: Abdul Rashid Bin Mohamed Shariff, -, Nazmi - Mat Nawi, -
Keywords: Oil palm ripeness, optical spectrometer, support vector machine, vegetation indices.
Abstract. Oil palm industry is the main source of income to Malaysia. Traditionally, oil palm fresh fruit bunches (FFB) are classified by oil palm fruit graders based on their experience. This method is subjective and often inconsistent. This paper reports a method to classify oil palm FFB using an optical spectrometer with wavelengths from 180 nm to 1100 nm. Ninety-six oil palm FFB were classified into unripe, ripe, and over ripe classes. Reflectance intensity data from 180 to 1100 nm were collected. Classification accuracies were tested by using all bands from 180 to 1100 nm, principal components, and seven vegetation indices (VI) that were computed from the extracted bands. The classifier applied in this study is Support Vector Machine. The results show that all bands classification achieved the highest accuracy of 90.6% with Kappa coefficient 0.86. Principal components classification shows 73.2% accuracy with 0.59 Kappa coefficient. Among the seven VIs, normalized difference vegetation index 2 (NDVI2) shows the highest accuracy of 84.4% with Kappa coefficient 0.77. These findings provide valuable information to future researchers in this field to develop automatic oil palm FFB classifier.
(Download PDF) (Export to EndNotes)