The development of data fusion technology calls for scientific, effective, objective evaluations of fusion systems. It is necessary to discuss the theories and methods for the performance evaluation (PE) of data fusion systems, in order to construct a general PE system for data fusion systems. As far as the characteristic level fusion and decision level fusion are concerned, a simulating environment is often established which yields the testing data, and the fusion algorithm can be tested and evaluated just like the black-box testing. In the paper, the performance evaluation methods of multisensor data fusion systems are discussed. A framework is presented for performance testing and evaluation of the multisensor data fusion. The common testing content and design method for the testing cases are discussed with the naval vessels data fusion system as the testing object. The calculating methods for the performance criterion of the fusion system are also presented.
[1]
Dale E. Blodgett,et al.
Analysis of the functional and real-time requirements of a Multi-Sensor Data Fusion (MSDF) / Situation and Threat Assessment (STA) / resource management (RM) system
,
1998
.
[2]
Vladimir S. Petrovic,et al.
Objective pixel-level image fusion performance measure
,
2000,
SPIE Defense + Commercial Sensing.
[3]
Dale E. Blodgett,et al.
Functional and real-time requirements of a multisensor data fusion (MSDF) situation and threat assessment (STA) resource management (RM) system
,
1998,
Defense, Security, and Sensing.