Scoring, Truthing And Registration Toolkit (START)

"AFRL currently has a large volume of imagery collected for evaluating Automatic Target Recognition (ATR) performance. However, this data cannot be used without adequate truthing information for each image. Truthing information typically consists of thetarget type and location of the targets present in the imagery. Specifying this information for a large number of images is tedious, time consuming, and error prone. In this Phase I proposal we present a complete truthing system we call the Scoring,Truthing, And Registration Toolkit (START). START uses the concept of a reference frame into which all imagery from a single data collection event (all image sequences of a target array from all sensors) is registered to reduce operator burnout associatedwith truthing a large number of images. The truth information is then specified once for the common reference frame. No additional information needs to be specified for the entire event, thereby greatly reducing the amount of effort expended truthing theimagery. START also provides a database engine to store all truthing, environment, and sensor information, enabling rapid construction of data subsets for testing ATR performance in specific operating conditions. Finally, START greatly simplifies ATRperformance evaluation by automating statistical scoring of all results. Tools for entering and maintaining truth information in imagery have potential uses in many government and industry domains