TY - GEN
T1 - Prediction system for soluble solid content in Averrhoa Carambola based on Vis-NIR image
AU - Rangkuti, Maisyarah Yuniar
AU - Saputro, Adhi Harmoko
AU - Imawan, Cuk
PY - 2018/1/9
Y1 - 2018/1/9
N2 - In this paper, a prediction system for a soluble solids content of Averrhoa Carambola or known as starfruit was developed using the Vis-NIR image. A push-broom hyperspectral imaging system is used to acquire Vis-NIR images from 200 sample of starfruit. All of the samples are prepared for the training (n=180) and testing (n=20) set over the range of 400-1000 nm. The hardware of the proposed system consists of a set of the workbench, controllable slider, two halogen light sources and a hyperspectral camera that is connected to PC via Camera Link. The software of system consists of reflectance image profile measurement, feature extraction, feature selection on spectral and spatial data, and soluble solids content prediction model. The partial least squares regression is used to build prediction models on full spectral data. The prediction model is used to get value prediction of soluble solids content. The predicted results compared with the reference measurement result of soluble solids content which obtained using a refractometer. The prediction model provided correlation coefficient of a testing set of 0.92 and root mean square errors of a testing set of 0.50. The results of our work indicate that there is a feasibility of implementing hyperspectral imaging technique on the nondestructive soluble solids content prediction of starfruit and suitable in an industrial sorting system for fruit quality.
AB - In this paper, a prediction system for a soluble solids content of Averrhoa Carambola or known as starfruit was developed using the Vis-NIR image. A push-broom hyperspectral imaging system is used to acquire Vis-NIR images from 200 sample of starfruit. All of the samples are prepared for the training (n=180) and testing (n=20) set over the range of 400-1000 nm. The hardware of the proposed system consists of a set of the workbench, controllable slider, two halogen light sources and a hyperspectral camera that is connected to PC via Camera Link. The software of system consists of reflectance image profile measurement, feature extraction, feature selection on spectral and spatial data, and soluble solids content prediction model. The partial least squares regression is used to build prediction models on full spectral data. The prediction model is used to get value prediction of soluble solids content. The predicted results compared with the reference measurement result of soluble solids content which obtained using a refractometer. The prediction model provided correlation coefficient of a testing set of 0.92 and root mean square errors of a testing set of 0.50. The results of our work indicate that there is a feasibility of implementing hyperspectral imaging technique on the nondestructive soluble solids content prediction of starfruit and suitable in an industrial sorting system for fruit quality.
KW - Averrhoa carambola
KW - hyperspectral
KW - image analysis
KW - partial least square regression
KW - soluble solids content
UR - http://www.scopus.com/inward/record.url?scp=85046014425&partnerID=8YFLogxK
U2 - 10.1109/ICELTICS.2017.8253269
DO - 10.1109/ICELTICS.2017.8253269
M3 - Conference contribution
T3 - Proceedings - 2017 International Conference on Electrical Engineering and Informatics: Advancing Knowledge, Research, and Technology for Humanity, ICELTICs 2017
SP - 114
EP - 118
BT - Proceedings - 2017 International Conference on Electrical Engineering and Informatics
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2017 International Conference on Electrical Engineering and Informatics, ICELTICs 2017
Y2 - 18 October 2017 through 20 October 2017
ER -