The paper proposes an improved version of Rol - based coding for compressing remote sensing images. A fully automatic background - clustering algorithm is used based on some features which have fully considered the characteristics of region of interest - based image compression scheme. The new model has three advantages: allowing near-lossless Rol segmentation; reducing the ambiguity of reconstructed Rol edge; enhancing intelligibility for Rol in reconstructed image. The paper also introduce Wold feature. The feature can capture random texture as well as structural texture, and hence is better than other purely stochastic texture features. Some results on remote sensing images arę presented.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.