 This paper presents a parallelized maximum likelihood classification approach for processing big imagery data using computer clusters and graphics processing units, which is scalable and suitable for change detection, object identification, and exploratory analysis in remote sensing applications. This article was authored by Shinsha and Bowishwa.