Avaliação de corpos de prova submetidos à ensaio de névoa salina utilizando o software Imagej para o processamento de imagens

Abstract

EVALUATION OF TEST SPECIMENS SUBJECTED TO SALT SPRAY TESTING USING IMAGEJ SOFTWARE FOR IMAGE PROCESSING. Corrosion assessment after salt spray tests has traditionally relied on visual inspection, which can limit the accuracy of quantitative evaluations. This study explores the application of image analysis through computational processing to measure the areas affected by corrosion on carbon steel plates, LN 38, coated with paints, subjected to salt spray testing conducted according to ASTM B117 standard. Using ImageJ software, we differentiate corrosion products from uncorroded metal surfaces based on variations in grayscale tones of the image converted to 8-bit and binarized, allowing precise measurement of corroded areas. Two 2² factorial experiments with replication were used in sample preparation, and each panel was evaluated following method 1 of ASTM D1654 standard, leading to the calculation of interaction effects with 95% confidence intervals for each condition. These computational results were then compared with evaluations made by an experienced technician through manual measurement of the panels. The findings suggest that computational image analysis can significantly improve the resolution and reliability of corrosion assessments compared to standard visual methods, indicating its potential as a valuable tool in corrosion research and evaluation.