复合马尔可夫模型在2——D谱估计中的应用

Spectrum Estimation of Textures with Complex Markov Random Field Model

马尔可夫模型做为纹理合成的重要手段之一已得到广泛地重视,复合马尔可夫模型是一种新的、绝对稳定的马尔可夫模型,其纹理合成视觉效果与一般马尔可夫模型相同。本文将之引入到2—D谱估计,实验结果表明其估值精度较高是纹理识别与理解的有力工具。

The Complex Marker Random Field model is an extension of the Markov Random Field Model. Its main advanage over the common Markov model is that a stable model always exist and is easy to estimte. It has found wide applications texture synthesis and compression. In this paper, it is adopted to do spectrum estimation for image textures. The experimental results show that it is a valid method for spectrum estimation and much more accurate than the common 2D-FFT spectrum estimation.