Comparative performance analysis of adaptive multispectral detectors

Target Identification and Detection Using LWIR Hyperpectral Signature Transformation of Multiple Missions without Registration

“…Strong signals from target types #1, #2, #3 are detected as well as false alarms from the clutter. Figure 6c shows output from the matched filter algorithm [18] applied to the "In-scene" image cube using the spectra from the transformed ("weighted CV" sampled with 1000 pixels or 10% coverage) Target #1 material. All objects associated with target type #1 are detected as well as a small signal from target type #3 as shown in Figure 6c.…”

Physics-Based Radiometric Signature Modeling and Detection Algorithms of Landmines Using Electro-Optical Sensors

“…A widely referenced approach to this problem is the RX algorithm presumably, named for its authors Reed and (Xiaoli) Yu [14]. This algorithm has been extended to include spectrally correlated (but spatially uncorrelated) clutter [15]. The spectral covariance matrix is estimated with an adaptive algorithm and used to design a spectral pre-prewhitening filter.…”

“…can exploit predicted signature information. The so-called "R'X" algorithm of Reed and Yu F14,15], which is based on a generalized likelihood ratio test (GLRT), is currently employed by the US Army and will be used as a baseline algorithm. RX makes relatively little use of mine signature information.…”

“…The RX algorithm [14,15] Performance evaluations of the. RX and EC algorithms are presented in Section 5.8 using a multispectral IR data set described in Section 2.7.2.…”

Kernel-Based Anomaly Detection in Hyperspectral Imagery

“…In spectral anomaly detection algorithms, pixels (materials) that have a significantly different spectral signature from their neighboring background clutter pixels are identified as spectral anomalies. Spectral anomaly detection algorithms [1][2][3][4][5] could also use spectral signatures to detect anomalies embedded within a background clutter with a very low signal-to-noise ratio. In spectral anomaly detectors, no prior knowledge of the target spectral signature are utilized or assumed.…”