Hyperspectral Data Processing: Algorithm Design and Analysis

29.4 Synthetic Image-Based Computer Simulations

In order to demonstrate the utility of proposed WSCA, experiments using computer simulations are conducted. For computer simulations, five Airborne Visible infra Red Imaging Spectrometer (AVIRIS) reflectance data, blackbrush, creosote leaves, drygrass, red soil, and sagebrush, are used and are shown in Figure 1.8. Each of these five spectral signatures has 158 bands after water bands are removed and can be considered as a 158-dimensional hyperspectral signature where each signature component is specified by a particular spectral wavelength.

Figure 29.8 Relationships derived from WSCA, which are WSCA-SST, WSCA-SSC, and signature self-discrimination, classification, and identification.

29.4.1 Signature Self-Tuning and Self-Denoising

Since blackbrush, creosote leaves, and sagebrush are very close in their spectral shapes, there are two cases for us to discuss: the first case is that one signature's details signatures are corrupted by any other signatures. The second case is that one of these three signatures are considered as a corrupted signature of the other two signatures, with both details and approximation signatures. Either case provides us with a good example to analyze WSCA in signature self-tuned performance. The experiments were divided into two categories. In the first one, the signature was self-tuned by detail signatures, ...