Feature-Level and Decision-Level Fusion of Noncoincidently Sampled Sensors for Land Mine Detection

We present and compare methods for feature-level (predetection) and decision-level (postdetection) fusion of multisensor data. This study emphasizes fusion techniques that are suitable for noncommensurate data sampled at noncoincident points. Decision-level fusion is most convenient for such data, but it is suboptimal in principle, since targets not detected by all sensors will not obtain the full benefits of fusion. A novel algorithm for feature-level fusion of noncommensurate, noncoincidently sampled data is described, in which a model is fitted to the sensor data and the model parameters are used as features. Formulations for both feature-level and decision-level fusion are described, along with some practical simplifications. A closed-form expression is available for feature-level fusion of normally distributed data and this expression is used with simulated data to study requirements for sample position accuracy in multisensor data. The performance of feature-level and decision-level fusion algorithms are compared for experimental data acquired by a metal detector, a ground-penetrating radar, and an infrared camera at a challenging test site containing surrogate mines. It is found that fusion of binary decisions does not perform significantly better than the best available sensor. The performance of feature-level fusion is significantly better than the individual sensors, as is decision-level fusion when detection confidence information is also available ("soft-decision" fusion).

[1]  Richard Lippmann,et al.  Neural Network Classifiers Estimate Bayesian a posteriori Probabilities , 1991, Neural Computation.

[2]  Y. Das,et al.  Analysis of an electromagnetic induction detector for real-time location of buried objects , 1990 .

[3]  John W. Wegrzyn,et al.  Sensor fusion performance gain for buried mine/UXO detection using GPR, EMI, and MAG sensors , 2000, Defense, Security, and Sensing.

[4]  Syama P. Chaudhuri,et al.  Multisensor data fusion for mine detection , 1990, Defense, Security, and Sensing.

[5]  Yogadhish Das,et al.  Multisensor vehicle-mounted teleoperated mine detector with data fusion , 1998, Defense, Security, and Sensing.

[6]  Belur V. Dasarathy,et al.  Decision fusion , 1994 .

[7]  Charles A. DiMarzio,et al.  Statistical fusion of GPR and EMI data , 1999, Defense, Security, and Sensing.

[8]  H. Vincent Poor,et al.  An Introduction to Signal Detection and Estimation , 1994, Springer Texts in Electrical Engineering.

[9]  S. Kay.  Fundamentals of statistical signal processing: estimation theory , 1993 .

[10]  H. Vincent Poor,et al.  An introduction to signal detection and estimation (2nd ed.) , 1994 .

[11]  Pramod K. Varshney,et al.  Distributed Detection and Data Fusion , 1996 .

[12]  D. L. Hall,et al.  Mathematical Techniques in Multisensor Data Fusion , 1992 .

[13]  Peter Howard,et al.  Performance results of the EG&G vehicle-mounted mine detector , 1999, Defense, Security, and Sensing.

[14]  Brian A. Baertlein,et al.  Subspace decomposition technique to improve GPR imaging of antipersonnel mines , 2000, Defense, Security, and Sensing.

[15]  A. Gunatilaka,et al.  Comparison of predetection and postdetection fusion for mine detection , 1999, Defense, Security, and Sensing.

[16]  Inder J. Gupta,et al.  Ramp response signatures for the detection of antipersonnel mines , 1999, Defense, Security, and Sensing.

[17]  Mahmood R. Azimi-Sadjadi,et al.  Detection of mines and minelike targets using principal component and neural-network methods , 1998, IEEE Trans. Neural Networks.

[18]  Klamer Schutte,et al.  Sensor fusion for antipersonnel landmine detection: a case study , 1999, Defense, Security, and Sensing.

[19]  Gregory A. Clark,et al.  Sensor feature fusion for detecting buried objects , 1993, Defense, Security, and Sensing.

[20]  Ibrahim K. Sendur,et al.  Techniques for improving buried mine detection in thermal IR imagery , 1999, Defense, Security, and Sensing.