Classification of Rice Varieties Using SIMCA Applied to NIR Spectroscopic Data.

Area of Science:

• Agricultural Science
• Analytical Chemistry
• Spectroscopy

Background:

• Accurate rice variety identification is crucial for achieving high and stable yields.
• Seed selection relies on the correct identification of rice varieties.
• Traditional methods can be time-consuming and less efficient.

Purpose of the Study:

• To explore the effectiveness of near-infrared spectroscopy (NIRS) combined with soft independent modeling of class analogy (SIMCA) for rapid rice variety identification.
• To establish and validate a classification model for distinguishing between different rice varieties.
• To provide a novel approach for accurate rice seed selection.

Main Methods:

• Utilized near-infrared spectroscopy (NIRS) for sample analysis.
• Applied soft independent modeling of class analogy (SIMCA) for classification modeling.
• Employed principal component analysis (PCA) as the basis for the SIMCA model.
• Validated the model using distinct modeling and test sets of rice samples.

Main Results:

• The SIMCA model achieved high accuracy in classifying rice varieties within the modeling set (100% for Kenjing No.5 and No.6, 97.5% for Kenjing No.9).
• The established model demonstrated 100% prediction accuracy for Kenjing No.5, Kenjing No.6, and Hongyu 001-1 in the test set.
• Kenjing No.9 samples showed 90% accuracy, with a 10% misidentification rate as Kenjing No.6.

Conclusions:

• The combination of NIRS and SIMCA provides an effective and rapid method for identifying rice varieties.
• This analytical approach offers a new and reliable strategy for the correct selection of rice planting varieties.
• The findings support the adoption of NIRS-SIMCA for quality control and varietal purity assessment in rice breeding programs.