Color PCA eigenimages and their application to compression and watermarking

摘要: From the birth of multi-spectral imaging techniques, there has been a tendency to consider and process this new type data as set parallel gray-scale images, instead an ensemble n-D realization. However, it proved that using vector-based tools leads more appropriate understanding color images thus efficient algorithms for processing them. Such are able take into consideration high correlation components successfully carry out energy compaction. In paper, novel method is proposed utilize principal component analysis in neighborhoods image order extract corresponding eigenimages. These eigenimages exhibit levels compaction such operations compression watermarking. Subsequently, two methods paper their comparison with available approaches presented.