Title :
Sensor registration using neural networks
Author :
Karniely, Haim ; Siegelmann, Hava T.
Author_Institution :
Dept. of Math. & Comput. Sci., Bar-Ilan Univ., Ramat-Gan, Israel
fDate :
1/1/2000 12:00:00 AM
Abstract :
One of the major problems in multiple sensor surveillance systems is inadequate sensor registration. We propose a new approach to sensor registration based on layered neural networks. The nonparametric nature of this approach enables many different kinds of sensor biases to be solved. As part of the implementation we develop some modifications to the common network training algorithm to tackle the inherent randomness in all components of the training set
Keywords :
feedforward neural nets; learning (artificial intelligence); radar tracking; sensor fusion; surveillance; target tracking; tracking filters; data association; gradient descent learning; inherent randomness; layered neural networks; multiple sensor surveillance systems; multiple target tracking; network training algorithm; radar sensors; sensor biases; sensor registration; supervised learning; systematic errors; target trajectories; track stability loss; Neural networks; Noise measurement; Particle measurements; Radar tracking; Sensor phenomena and characterization; Sensor systems; State estimation; Surveillance; Target tracking; Trajectory;
Journal_Title :
Aerospace and Electronic Systems, IEEE Transactions on
