| Autor: |
Shiddiq, Minarni, Herman, Arief, Dodi Sofyan, Yesyaf, Oktaviandri Prima, Lase, Chrisvon, Husein, Ikhsan Rahman, Ayunda, Sri |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 9/20/2022, Vol. 2663 Issue 1, p1-9, 9p |
| Abstrakt: |
Volume measurement has used computer vision to estimate the mass of many types of fruits. Capacity determination of a crude palm oil refinery involves the measurement of weight and mass of oil palm fresh fruit bunches, which is usually measured using a weighbridge. Automation is possible using a fast sorting and grading machine vision where computer vision is applied. This study aimed to use a computer vision method with a cross-line laser beam to measure the volumes of oil palm FFB using a segmentation process and an ellipsoid equation. Samples were 40 oil palm FFB of Tenera Variety. The volumes measured using two other methods, the water displacement method (WDM) and the manual method, were used for validation. The average difference of the oil palm FFB volumes measured manually and by computer vision was 0.77 x 103 cm3 with R2 of 0.9336. The volumes measured using the WDM method and computer vision has an average difference of 0.81 x 103 cm3 with an R2 of 0.9074. Comparison of manual to computer vision method and WDM to computer vision method show the volumes measured are within 95 % spread using Bland-Altman plot. The FFB volumes measured using the computer vision method range from 3.92×103 cm3−18.02×103 cm3 which relate to 2.4 - 17.9 kg of masses. These results show the potential use of the computer vision method for oil palm FFB. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
