为了在抑制噪声的同时保护图像的细节,提出了一种基于模糊函数的自适应平滑约束图像复原算法。该算法首先用模糊函数对图像局部区域内线元素的数量和方向进行评价,然后根据评价结果使用不同的高通滤波器构造平滑约束条件,同时使平滑约束条件随着图像复原的迭代过程不断更新,以便自适应地在图像平坦区域抑制噪声,而在存在物体边界的区域则保护细节信息。对比实验表明,此方法具有更高的收敛速度,更好的客观评价指标和主观视觉效果。

A new degraded image restoration algorithm which uses adaptive smoothness constraint based on fuzzy functions is used,is proposed.In order to suppress noise in image,simultaneously,preserve the details during image restoration,the algorithm uses the fuzzy functions to estimate the number and orientation of line elements in each image region,then these edge estimation results lead to Corresponding smoothness constraint.Furthermore,the smoothness constraint is updated along with image restoration iteration,such that the algorithm can suppress noise in flat regions and preserve details of existing edge.The experimental results show that it has faster convergence speed,better objective quality and subjective vision effect.