Effects of using enhanced input range profiles for 1-d automated maritime vessel classification

This research deals with quantifying the benefits of using the novel Sum Normalized Range Profile (SNRP) over prior art for 1-d feature extraction based classification of maritime vessels. For a fair comparison, classification and feature extraction techniques are maintained the same, for each range profile input. While utilizing SNRP inputs, the automated target recognition engine results indicate an improvement of roughly 8.2% and 11.3% for the correct probability of classification and joint-classification, respectively. This finding has ramifications to enhance not only 1-d classification techniques but also data fusion techniques that can encompass multidimensional and multiple datasets.