Near infrared diffuse reflectance measurement set-up in non-destructive brix measurement for intact pineapples
Conventional Brix prediction of pineapple is time-consuming and destructively. Even few of the applications of near infrared spectroscopy in diffuse reflection mode had been implemented in determining the Brix of pineapple, those applications did not investigate the effect of pineapple surface (geom...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English English |
| Published: |
2021
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/8264/1/24p%20HONG%20FAN%20WEI.pdf http://eprints.uthm.edu.my/8264/2/HONG%20FAN%20WEI%20COPYRIGHT%20DECLARATION.pdf http://eprints.uthm.edu.my/8264/3/HONG%20FAN%20WEI%20WATERMARK.pdf http://eprints.uthm.edu.my/8264/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Conventional Brix prediction of pineapple is time-consuming and destructively. Even few of the applications of near infrared spectroscopy in diffuse reflection mode had been implemented in determining the Brix of pineapple, those applications did not investigate the effect of pineapple surface (geometrical effect) and the measurement setup on the prediction performance. This study investigates the geometrical effect with three parameters including the effect of different parts of pineapple on the diffuse reflectance, relationship between diffuse reflectance and soluble solids content (SSC), and effect of measurement position on the calibration model. The measurement setup was investigated through pre- and post-dispersive (with fiber optic design) NIRS technique. Three parameters were studied with pre-dispersive sensor. First, the effect of acquired difference reflectance from different parts of pineapple i.e. top, middle, and bottom was investigated with boxplot technique. Then, the relationship between the diffuse reflectance and SSC was evaluated with correlation plot. Lastly, nine independent artificial neural networks were separately trained to investigate the geometrical effects on different parts of pineapple. Results show that the concave surface of top and bottom parts of pineapples would affect the reflectance of light and deteriorate the prediction performance. With bifurcated fiber optic design, the predictive model of middle part of pineapples achieved the best performance i.e. root mean square error of prediction and correlation coefficient of prediction of 1.313 °Brix and 0.7408 respectively. The main finding of the research is that the geometrical effect that affects the Brix prediction can be minimized by proper measurement setup. |
|---|