Hyperspectral imaging has become a standard for most applications that require precision based analysis. This is due to the fine spectral resolution that hyperspectral data offers. Target detection based on hyperspectral imaging is one of the significant applications required for numerous defence, surveillance as well as many civilian studies. This involves detailed analysis of all bands of hyperspectral data for presence of the desired target. However, there exist many parameters which may have bearing on the performance of detection algorithm. These include sensor related parameters like noise, calibration etc., spatial parameters like size, shape and location etc. and scene parameters like illumination variation, target composition, colour, background etc. This paper demonstrates the implications of three scene parameters namely illumination, background and colour on detection of many known targets using the hyperspectral data. The hyperspectral data acquired over Rochester Institute of Technology (RIT) for experimental purposes has been used. Three popular detection algorithms namely, ACE, MF, SAM have been implemented for target detection and the impact of selected parameters is assessed.
Hyperspectral imaging is a powerful tool in the field of remote sensing and has been used for many applications like mineral detection, detection of landmines, target detection etc. Major issues in target detection using HSI are spectral variability, noise, small size of the target, huge data dimensions, high computation cost, complex backgrounds etc. Many of the popular detection algorithms do not work for difficult targets like small, camouflaged etc. and may result in high false alarms. Thus, target/background discrimination is a key issue and therefore analyzing target’s behaviour in realistic environments is crucial for the accurate interpretation of hyperspectral imagery. Use of standard libraries for studying target’s spectral behaviour has limitation that targets are measured in different environmental conditions than application. This study uses the spectral data of the same target which is used during collection of the HSI image. This paper analyze spectrums of targets in a way that each target can be spectrally distinguished from a mixture of spectral data. Artificial neural network (ANN) has been used to identify the spectral range for reducing data and further its efficacy for improving target detection is verified. The results of ANN proposes discriminating band range for targets; these ranges were further used to perform target detection using four popular spectral matching target detection algorithm. Further, the results of algorithms were analyzed using ROC curves to evaluate the effectiveness of the ranges suggested by ANN over full spectrum for detection of desired targets. In addition, comparative assessment of algorithms is also performed using ROC.
Most military targets of strategic importance are very small in size. Though some of them may get spatially resolved, most cannot be detected due to lack of adequate spectral resolution. Hyperspectral data, acquired over hundreds of narrow contiguous wavelength bands, are extremely suitable for most military target detection applications. Target detection, however, still remains complicated due to a host of other issues. These include, first, the heavy volume of hyperspectral data, which leads to computational complexities; second, most materials in nature exhibit spectral variability and remain unpredictable; and third, most target detection algorithms are based on spectral modeling and availability of a priori target spectra is an essential requirement, a condition difficult to meet in practice. Independent component analysis (ICA) is a new evolving technique that aims at finding components that are statistically independent or as independent as possible. It does not have any requirement of a priori availability of target spectra and is an attractive alternative. This paper, presents a study of military target detection using four spectral matching algorithms, namely, orthogonal subspace projection (OSP), constrained energy minimisation, spectral angle mapper and spectral correlation mapper, four anomaly detection algorithms, namely, OSP anomaly detector (OSPAD), Reed–Xiaoli anomaly detector (RXD), uniform target detector (UTD), a combination of RXD–UTD. The performances of these spectrally modeled algorithms are then also compared with ICA using receiver operating characteristic analysis. The superior performance of ICA indicates that it may be considered a viable alternative for military target detection.