Random Forest Algorithm Based Strain Analysis on Composite Materials using Digital Image Processing

Abstract

For engineers, knowledge of strain analysis is crucial as it has a critical role in several engineering designs and practical work. Condition monitoring of various composite structures is very crucial in engineering in the wind, civil, aviation, and auto sectors. This article provides comparison study of strain analysis of jute material using conventional method, Digital image correlation, regression analysis and random forest algorithm. The result shows that mean square error obtained using random forest algorithm is 3.8*10^-4 which is least in comparison to contact- basedmethod.