摘要
针对H1模型和TV模型在图像恢复方面的不足,提出一种改进的自适应TV去噪模型。该模型能在边缘附近自动地选择保边较好的TV去噪模型,而在远离边缘处自动选择平滑模型,并且同时对点扩散函数(PSF)进行约束。采用频域交替迭代的方法,在恢复出图像的同时也可将PSF恢复出来。MATLAB仿真结果表明,采用的方法在去噪的同时很好地保留了图像边缘和纹理信息,避免了阶梯效应,复原图像的峰值信噪比(PSNR)与其它方法相比有很好的提升。
On account of the deficiencies of H1 model and the TV model in image restoration,an improved adaptive TV denoising model was presented in this paper.The model can automatically select the denoising TV model near the edges to keep edge better,and select a smooth model away from the edges,meanwhile making constraints to the point spread function（PSF）.Alternating iterative method in frequency domain is adopted,and both the image and the PSF can be recovered simultaneously.Simulation results in MATLAB show that the method can keep the image edge and texture information while denoising,avoiding the staircase effect.Peak signal to noise ratio（PSNR） of the restored image is greatly improved compared with other methods.
出处
《长春理工大学学报（自然科学版）》
2010年第4期98-100,共3页
Journal of Changchun University of Science and Technology(Natural Science Edition)
