A comparative study of target detection algorithms for hyperspectral imagery

In this paper, we review and compare the state-of-the-art target detection algorithms. We introduce a new target detection workflow incorporating a Minimum Noise Fraction (MNF) transform before target detection. Applying a MNF transform was found to improve the detection results in general, especially with the Orthogonal Subspace Projection detector. In this paper, we propose a new algorithm - Mixture Tuned Target-Constrained Interference-Minimized Filter (MTTCIMF). MTTCIMF uses the MNF transformed image as the input and combines the mixture tuned technique with the TCIMF target detector. By adding an additional infeasibility band, mixture tuned techniques improve the detection results with a reduced number of false alarms. A HyMap data set with ground truth is used in the comparative study. Quantitative and visual evaluation of different algorithms is given. A new quantitative metric is proposed to evaluate the visibility of targets in the detection results. Keywords: target detection, mixture tuned matched filter (MTMF), mixture tuned target-constrained interferenceminimized filter (MTTCIMF), minimum noise fraction (MNF), adaptive coherence estimator (ACE), orthogonal subspace projection (OSP), constrained energy minimization (CEM), target visibility

[1]  J. Boardman,et al.  Leveraging the High Dimensionality of AVIRIS Data for improved Sub-Pixel Target i Unmixing and Rejection of False Positives : Mixture Tuned Matched Filtering , 1998 .

[2]  Chein-I Chang,et al.  Generalized constrained energy minimization approach to subpixel target detection for multispectral imagery , 2000 .

[3]  Dimitris G. Manolakis,et al.  Detection algorithms for hyperspectral imaging applications , 2002, IEEE Signal Process. Mag..

[4]  Anthony J. Ratkowski,et al.  The sequential maximum angle convex cone (SMACC) endmember model , 2004, SPIE Defense + Commercial Sensing.

[5]  Jason E. West,et al.  Matched Filter Stochastic Background Characterization for Hyperspectral Target Detection , 2005 .

[6]  Louis L. Scharf,et al.  The adaptive coherence estimator: a uniformly most-powerful-invariant adaptive detection statistic , 2005, IEEE Transactions on Signal Processing.

[7]  Chein-I Chang,et al.  Further results on relationship between spectral unmixing and subspace projection , 1998, IEEE Trans. Geosci. Remote. Sens..

[8]  Steven Kay,et al.  Fundamentals Of Statistical Signal Processing , 2001 .

[9]  Steven Johnson.  Constrained energy minimization and the target-constrained interference-minimized filter , 2003 .

[10]  Chein-I Chang,et al.  Hyperspectral image classification and dimensionality reduction: an orthogonal subspace projection approach , 1994, IEEE Trans. Geosci. Remote. Sens..

[11]  Gary A. Shaw,et al.  Hyperspectral Image Processing for Automatic Target Detection Applications , 2003 .

[12]  John P. Kerekes,et al.  Development of a Web-Based Application to Evaluate Target Finding Algorithms , 2008, IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium.

[13]  Chein-I Chang,et al.  Target-constrained interference-minimized approach to subpixel target detection for hyperspectral images , 2000 .

[14]  Chein-I. Chang.  Hyperspectral Imaging: Techniques for Spectral Detection and Classification , 2003 .