A new fast search algorithm for image vector quantization

In image vector quantization (VQ), the computing complexity of the classical full search algorithm increases exponentially with vector dimension and the quantization bit rate. A new fast search algorithm using energy band segmentation (EBS) on the basis of examining the energy property between the input vector and codevector of a remote sensing image (RSI) is presented. Experimental results show that this method could decrease the search scope and speed up the encoding process on condition that the reconstructed image quality is hardly sacrificed. Meanwhile, this method has an advantage over other fast algorithms in terms of hardware realization