Automated Missile Aim Point Selection Technology.
Abstract
Martin Marietta Labs - Baltimore (MMLB) has developed missile aimpoint algorithms that address the requirements of the ONR AMAPS program by maintaining an awareness of the dominant issues. To be of practical use to the AMAPS program, Automatic Target Recognition (ATR) algorithms must be effective and reliable over a wide range of weather, lighting, visibility, obscuration, and terrain conditions in cluttered and noisy environments. There should be minimal restrictions on type of target, target orientation, and viewing direction. Further, effective performance must be achieved within the power, speed, size, and cost constraints dictated by real world operational conditions. The model-based approach to ATR (MBATR) has shown much promise in achieving high levels of target discrimination of reduced target signatures in highly cluttered environments (as required for target identification sufficient for distinguishing between two similar targets in the same class i.e., friendly tank from enemy tank in battlefield conditions). The model based approach derives target information from target and sensor models, so the target information is not limited by an unwieldy training set of images and the tedious and unreliable training task is not required. The model. based approach achieves robustness of recognition by finding an optimum match between target models and image data through a powerful hypothesis prediction and verification scheme, (see figure 1).
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1995
- Accession Number
- ADA300813
Entities
People
- Gary Whitten
Organizations
- Martin Marietta