You are here

RF-IR Data Fusion for Track and Data Correlation


OBJECTIVE: To develop new methods for multi-sensor target handover and characterization. DESCRIPTION: For radars fielded by the Missile Defense System, there is a given set of available features that were developed for acquisition, track and discrimination of targets. Similarly, for electro-optic/infrared sensors there are standard features which have been developed for acquisition, track and discrimination. These features may not be the optimal features for target correlation between sensors. Thus, this task seeks the development of either innovative new features to aid multi-sensor track correlation, or innovative data fusion techniques with the existing features to aid correlation. Additionally, realizing that the RF features for discrimination exploit a given set of target attributes, and similarly for the EOIR, we seek innovative data fusion processes to better characterize the target. The first subtask, correlation of objects between sensors, is a very important handover function in multi-sensor, multi-target data fusion, particularly when sensors are sparse and overlapping coverage may not occur. This can become complicated when different types of sensor systems (RF versus EOIR) are viewing the same objects. Success in this subtask will be a demonstration of improved handover, compared with a metric only procedure, from EO/IR to RF sensors and from RF to EO/IR, when multiple objects are in the respective scenes. Appropriate sensor geometries with respect to target will need to be considered. The second, or alternate, subtask will be the development of data fusion techniques to combine information from the disparate sensors - to understand the appropriate contributions to target identity from the various sources. Object classification schemes may be considered, however, they must be robust to objects not in the training set, and various subsets of available features. Other techniques, such as Bayes net approaches or influence diagrams, are also welcome to explore innovative, robust methods for target characterization from multi-sensor features. Success will be a demonstration of target characterization in RF, in EO/IR, then together, to demonstrate how a target can be more accurately characterized using a combination of RF and EO/IR measurements. Previous efforts have not dealt with the long ranges and specific available features for the MDA engagement for data fusion, and correlation is a major outstanding challenge. EOIR sensors are relatively new to the sensor suite for MDA, as well, and present new challenges for exploitation. The researcher may include multiple EO/IR bands as well as several radar frequencies, if desired in their analyses. However, targets will be at ranges that will cause them to appear on, at most, one pixel for the EO/IR focal plane. PHASE I: Develop and demonstrate, through proof-of-principle tests, target correlation or characterization improvements using measurement data from disparate sensors. PHASE II: Refine and update concept(s) based on Phase I results and demonstrate the technology in a realistic environment using agency provided engagements. Demonstrate the technology"s ability real-time in a stressed environment; with few sensors, and many targets. PHASE III: Demonstrate the new technologies via operation as part of a complete system or operation in a system-level test bed to allow for testing and evaluation in realistic scenarios. Market technologies developed under this solicitation to relevant missile defense elements directly, or transition them through vendors. COMMERICALIZATION: The contractor will pursue commercialization of the various technologies and optimization components developed in Phase II for potential commercial and military uses in many areas such as automated processing, robotics, medical industry, and in manufacturing processes.
US Flag An Official Website of the United States Government