A SEDEA-based approach for evaluating the effectiveness of thermal camouflage using image and spectral data

Abstract

Using military camouflage technologies in modern warfare to avoid or reduce the ability of detection from opponents has rapidly developed all over the world. The method to evaluate camouflage effectiveness is investigated to ensure the practical applicability of camouflage products. Because of sensitive information, no universally accepted or reliable standards exist for measuring camouflage effectiveness domestically or internationally. This absence of standards has led to various assessment methods, broadly categorized into subjective and objective evaluations. Each approach has certain advantages and restrictions, with neither proving unequivocally superior. A comprehensive evaluation of camouflage effectiveness is a complex issue that requires considering multiple critical factors such as the characteristics of surveillance equipment, targets, background environment, atmosphere conditions, and observers. This study introduces a novel objective evaluation framework utilizing the new Data Envelopment Analysis (DEA) model, incorporating a newly developed set of assessment indicators derived from the target’s thermal image data and radiometric spectral data. This model utilizes 11 evaluation indicators, including eight characteristic indices derived from thermal images captured by a thermal camera and three spectral feature indices obtained from data collected by a spectroradiometer. Experimental analysis on multiple camouflage outfit samples demonstrate that, compared to traditional DEA models, the Super-Efficiency DEA model offers a more reliable ranking system, effectively categorizing the effectiveness of camouflage outfits. The results also suggest that the Super-Efficiency DEA model can serve as a valuable tool for standardizing and enhancing the objectivity of camouflage effectiveness assessments in the future.