A new approach to VQ-based compression and classification of sensor data

The size and power resource constraints associated with the airborne platforms of various sensor systems require that sensor data be transmitted to ground based automatic target recognition (ATR) processing stations. However, the bandwidth limitations of feasible communication channels restrict the amount of data that can be reliably transmitted. This presentation gives an overview of a new technology that permits efficient clipping service on board the airframe that is based on the use of vector quantization (VQ). The new technology generates a mathematical decomposition of sensor data called "direct sum successive approximations" (DSSA). DSSA technologies permit low-level ATR functions to be combined with VQ data compression functions to form an intelligent clipping service.