New Method for Adaptive Karhunen-Loeve Color Transform

In the paper is offered one new method for color space reduction, based on the Karhunen-Loeve Transform (KLT). The use of the KLT for the processing of the image primary color components gives as a result higher decorrelation, which ensures the enhancement of other operations, such as: compression, color-based segmentation, etc. In order to reduce the computational complexity of KLT is assumed that the color vectors describe a stationary-random process (first order Markov process). Depending on the relations obtained, are calculated the corresponding KLT matrices for the color vectors transform. The presented approach is based on the analytic determination of the color covariance matrix eigenvectors. The new algorithm surpasses the existing similar algorithms in its lower computational complexity, which is a prerequisite for fast color segmentation or for adaptive color image coding, aimed at real time applications.