Solar modules are the main component of renewable energy systems that convert
sunlight into electricity. Over time, the surface of the modules can suffer damage such as cracks,
scratches, and stains that reduce efficiency and power output. Manual inspection has
limitations in terms of time, cost, and potential for error, so a reliable automatic detection
method is needed. Advances in machine learning are being utilized to model complex patterns,
particularly through texture-based image analysis that represents spatial correlations between
pixels. This study uses the Gray Level Co-occurrence Matrix and Local Binary Pattern methods
for texture feature extraction. GLCM is calculated at four angle orientations (0°, 45°, 90°, 135°)
with contrast, dissimilarity, homogeneity, energy, ASM, and damage percentage parameters.
LBP uses the number of neighbors and radius parameters, with mean, variance, and entropy
features. Damage segmentation was performed using Otsu Thresholding to determine the
optimal threshold and calculate the area of damage. The results show that GLCM achieved
stable accuracy of 100% in all experiments, with the highest damage area of 35%. The LBP
method achieved 67% accuracy with a maximum damage percentage of 27%. These findings
indicate that GLCM is more effective in class separation, while LBP is capable of capturing local
texture patterns. This model has the potential to support automatic solar module maintenance
and improve solar energy utilization efficiency.