Application of laser-induced backscattering imaging system for classifying different ripening stages of Musa Acuminata cv. Berangan bananas
Inadequacy and inefficiency of monitoring quality systems for fruits have made a great impact that leads to an increasing number of post-harvest losses as they could have been damage during storage. Fruits undergone complex changes in their biochemical and physicochemical during ripening process....
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2017
|
| Subjects: | |
| Online Access: | http://psasir.upm.edu.my/id/eprint/71190/1/FK%202017%2060%20-%20IR.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Inadequacy and inefficiency of monitoring quality systems for fruits have made a great
impact that leads to an increasing number of post-harvest losses as they could have been
damage during storage. Fruits undergone complex changes in their biochemical and
physicochemical during ripening process. This study evaluates the potential of the
backscattering imaging system to evaluate the bananas at different ripening stages.
Backscattering image (BSI) of Musa Acuminata cv. Berangan was captured by a charge
coupled device (CCD) camera and a laser diode emitting light at 658 nm. The system
consisted of CCD camera with a zoom lens (focal length 18-108mm), a solid state laser
diode of 658 nm at 1mm diameter as a light source and a computer equipped with an
image processing software for automated image analysis. A total number of 360 samples
of Musa Acuminata cv. Berangan from ripening stages 2 to 7 with 60 samples per stage
group were used in this study. The gray level intensity and size of the backscattering area
were used for estimating the quality properties of bananas. The results showed that the
highest correlation was found between BSI parameters and total soluble solids content
(TSS). Moreover, linear discriminant models were built for the two- class (unripe, ripe)
and six-class (based on the commercial colour index) of ripening stages classifications.
The overall accuracy for two-class and six-class classifications resulted in 94.2% and
59.2% classification accuracies, respectively. It can be concluded that the laser lightinduced
backscattering imaging could be potentially used for predicting the ripening
stages of bananas and could be further developed for an automated quality control
system. |
|---|