摘要:
本文介紹了一種實際散斑模式的數(shù)學模型和噪聲統(tǒng)計模型,并提出了一種針對這種模型的自適應次優(yōu)濾波方法。文中在分析了散斑模式及其噪聲性質的基礎上,利用其局部方向性特征,結合最優(yōu)線性濾波器和非線性濾波器的特點,對線性最小均方誤差濾波器進行了自適應逼近。實驗結果表明,對散斑模式而言,本文的濾波方法與其它常用的圖象濾波方法相比,具有更好的去噪和邊緣保護性能,并且具有較好的濾波韌性。
Abstract:
A statistical noise model and a mathematical model for real speckle pattern are presented, and then, in view of the models, a new adaptive suboptimal image filtering approach is proposed. The proposed approach, with the local direction features of speckle pattern, combines the characteristcs of optimal linear filter with non-linear filter and is an adaptive approximation to linear minimum mean square error filter. Experimental results show that the proposed approach has fairly good edge-preserved performance, compared with other present image filters, as well
as much better filtering performance and robustness for speckle pattern.
下載:
