Deteksi Keaslian Beras Aceh Varietas Sigupai Menggunakan Portable Near Infrared Spectrometer
Abstract
Sigupai merupakan varietas beras aromatik yang menjadi ciri khas kabupaten Aceh Barat Daya. Banyaknya varietas beras dengan kemiripan karakter, memperbesar peluang pencampuran beras dalam pendistribusian. Tujuan dari penelitian ini adalah memprediksi keaslian beras sigupai dan persentase tingkat keaslian beras sigupai secara non-destruktif menggunakan portable near-infrared (NIR) spectrometer. Penelitian ini dianalisis dengan metode Partial Least Square-Driscriminant Analysis (PLS-DA) dan Partial Least Squares-Regression (PLS-R) pada dua perlakuan. Perlakuan pertama, beras yang digunakan adalah sigupai murni (tanpa campuran) dan sigupai campuran (sigupai dan inpari dengan persentase 1% sampai 30% atau pada tingkat keaslian 69,9% sampai 99%) masing-masing 43 sampel. Perlakuan kedua, beras yang digunakan adalah sigupai yang dicampur dengan inpari pada konsentrasi 0% sampai 30% atau pada tingkat keaslian 69,9% sampai 100% sebanyak 44 sampel. NeoSpectra dan SCiO dengan metode Partial Least Square-Driscriminant Analysis (PLS-DA) mampu memprediksi keaslian beras sigupai dengan hasil terbaik menggunakan data original (tanpa pre-treatment). Hasil terbaik dengan NeoSpectra menghasilkan nilai akurasi, sensitivitas, spesifisitas, dan false positive rate, yaitu masing-masing 89,29%; 92,86%; 85,71% dan 14,29%, sedangkan SCiO menghasilkan nilai akurasi, sensitivitas, spesifisitas, dan false positive rate, yaitu masing-masing 97,44%; 100%; 94,87% dan 5,13%. Analisis dengan metode Partial Least Squares-Regression (PLS-R) juga mampu menduga persentase tingkat keaslian beras sigupai dengan hasil terbaik menggunakan pre-treatment derivative SG1. NeoSpectra menghasilkan nilai koefisien korelasi kalibrasi, koefisien korelasi validasi, SEC, SEP, RPD, dan konsistensi, yaitu masing-masing 0,99; 0,96; 1,52%; 1,50%; 5,93, dan 100,69%. SCiO menghasilkan nilai koefisien korelasi kalibrasi, koefisien korelasi validasi, SEC, SEP, RPD, dan konsistensi, yaitu masing-masing 0,99; 0,98; 1,37%; 1,31%; 6,83 dan 104,78%. Hasil penelitian menunjukkan bahwa portable near-infrared spectrometer, baik NeoSpectra maupun SCiO berpotensi digunakan sebagai alat analisis keaslian beras sigupai di lapangan. Sigupai is one of the local rice varieties that characterizes the Southwest Aceh district. Wide rice varieties with similar characters increase the chances of mixing rice in distribution.The purpose of this study was to non-destructively predict the authenticity of sigupai rice and the percentage of authenticity in sigupai rice using a portable near-infrared (NIR) spectrometer. This study was analyzed using the Partial Least Square-Driscriminant Analysis (PLS-DA) and Partial Least Squares-Regression (PLS-R) methods for the two treatments. In the first treatment, the rice used was pure sigupai (without mixture) and mixed sigupai (sigupai and inpari with a percentage of 1% to 30% or at an authenticity level of 69,9% to 99%), each of 43 samples. In the second treatment, the rice used was sigupai mixed with inpari at a concentration of 0% to 30% or at an original level of 69,9% to 100% for 44 samples. NeoSpectra and SCiO with the Partial Least Square-Driscriminant Analysis (PLS-DA) method were able to predict the authenticity of sigupai rice with the best results using original data (without pre-treatment). The best results with NeoSpectra produced accuracy values, sensitivity, specificity and false positive rate values, namely 89,29%; 92,86%; 85,71% and 14,29%, while SCiO produced accuracy values, sensitivity, specificity, and false positive rate, namely 97,44%; 100%; 94,87% and 5,13%. Analysis using the Partial Least Squares-Regression (PLS-R) method was also able to estimate the percentage of authenticity in sigupai rice with the best results using pre-treatment derivatives SG1. NeoSpectra produces values of a calibration correlation coefficient, validation correlation coefficient, SEC, SEP, RPD, and consistency, namely 0,99; 0,96; 1,52%; 1,50%; 5,93 and 100,69%. SCiO produces values of a calibration correlation coefficient, validation correlation coefficient, SEC, SEP, RPD, and consistency, namely 0,99; 0,98; 1,37%; 1,31%; 6,83 and 104,78%. The results showed that the portable near-infrared spectrometer, both NeoSpectra and SCiO, has the potential to be used as an instrument for analyzing the authenticity of sigupai rice in the field.
Collections
- MT - Agriculture Technology [2270]