Real-time ordnance recognition in color imagery

Military test ranges containing unexploded ordnance due to five-fire testing and training exercises are a significant safety problem in many locations. Automated cleanup of such sites is a challenging goal. Towards this end, this paper describes algorithms to detect a type of ordnance in current usage using color imagery. The techniques are designed to run in real-time using off-the-shelf hardware for use on an unmanned ground vehicle. Our methodology is to quickly detect candidate locations using color and stereo information. Additional processing is Men applied to the candidate locations in order to eliminate false alarms. This technique has been tested on a set of imagery from a live-fire test range with outstanding results.