Image Enhancement for Graininess Quantification in Metallic Coatings: The Role of Image Calibration

Graininess in metallic coatings is a critical factor influencing visual quality, particularly in automotive coatings. This study aims to quantitatively assess graininess in metallic automotive coatings and evaluate the effect of image calibration as a preprocessing step on its correlation with human visual perception. A total of ۳۰ metallic coatings were analyzed using two methods: Power Spectral Density (PSD) derived from the Fourier Transform and features extracted from the Gray Level Co-occurrence Matrix (GLCM). The PSD results from calibrated images showed a higher Pearson correlation coefficient with visual assessments compared to raw images (۰.۸ vs. ۰.۶), indicating the sensitivity of the Fourier Transform to image preprocessing. GLCM-based features, such as contrast, correlation, and homogeneity, demonstrated R² > ۰.۸ for both raw and calibrated images, indicating their reliability across different conditions. However, the entropy feature significantly improved after calibration, with its correlation increasing from ۰.۶ to ۰.۸, highlighting its sensitivity to preprocessing.

کلیدواژه ها:

Metallic Coatings ، Graininess ، Visual Experiment ، GLCM ، Fourier Transform