<p>Comprehensive Soil Image Features Extraction Using Color, Luminance, and Gray-Level Co-Occurrence Matrix Analysis</p>

This paper presents a comprehensive framework for soil image feature extraction and classification using RGB color, luminance, and gray-level co-occurrence matrix (GLCM)-based texture descriptors. A dataset of 1,300 soil images representing black, laterite, red, and white soil categories was used for evaluation. The proposed framework integrates color, brightness, and texture information to improve soil representation and classification performance. K-Nearest Neighbors classification was evaluated using Euclidean, Manhattan, Chebyshev, and Minkowski distance metrics. Experimental results demonstrate that feature fusion improves classification performance compared to individual feature domains. Among the evaluated metrics, Euclidean and Chebyshev distances achieved strong classification performance across different feature sets. The proposed framework provides a scalable and efficient solution for image-based soil analysis and precision agriculture applications. Future work will focus on larger and more diverse datasets for real-world validation.