<![CDATA[Tea leaf quality serves as the fundamental determinant of both sensory characteristics and commercial competitiveness in the global tea market. However, manual assessment of tea leaf quality remains limited by observer subjectivity and inconsistent classification. This study aims to develop an automatic tea leaf quality classification system based on leaf maturity using a digital image processing approach.The method employed is Support Vector Machine (SVM) with a combination of three feature extraction techniques: color histogram for color features, Gabor filter for texture features, and Histogram of Oriented Gradients (HOG) for shape features. The dataset consists of 4,272 tea leaf images classified into four quality classes: Premium Grade, Standard Grade, Basic Grade, and Reject Grade. Principal Component Analysis (PCA) was applied for dimensionality reduction while maintaining 95% data variance. Testing results show an accuracy of 84.53% with an F1-score of 84.56%, demonstrating the effectiveness of the system in automatically classifying tea leaf quality]]>
Copyrights © 2026
