多小波描述的通用隐马尔可夫树模型图像去噪算法

小波域通用隐马尔可夫树(uHMT)模型充分利用了实际图像内部的自相似性,仅用9个参数(与图像的大小和小波的尺度数目无关)就可以完全确定实际图像的隐马尔可夫树(HMT)模型,极大地简化了隐马尔可夫树模型,但这使得图像去噪的精度降低。多小波描述在图像去噪方面取得了较好的效果。利用通用隐马尔可夫树(uHMT)模型和多小波描述各自在图像去噪方面的优势,将两者结合起来,提出了一种称之为多小波描述的通用隐马尔可夫树模型(M-uHMT)的新的图像去噪算法。仿真结果表明,这种算法的去噪效果优于典型的去噪算法。

Image De-noising Algorithm Using Multiple Wavelet and Universal Hidden Markov Tree Model

Wavelet-domain universal hidden Markov tree(uHMT) simplify the hidden Markov tree(HMT) model to specify it with just only mine parameters(independent of the size of the image and the number of wavelet scales) by exploiting the inherent self-similarity of real-world images,but it makes image denoising become less accurate.Multiple wavelet representations have excellent performance in image de-noising.In this paper,combining the multiple wavelet representations with the universal hidden Markov tree and using their advantages in image de-noising,we propose a new image de-noising algorithm,called M-uHMT.It is simple and effective.Simulation results show that the proposed M-uHMT can achieve the state-of-the-art image de-noising performance at low computational complexity.