摘要
针对调幅加网印刷品扫描图像,提出一种自适应去网纹算法。该算法根据CMYK四色印刷规律及彩色图像上影响图像质量比较显著的摩尔网纹形成的原因,通过在傅立叶频谱(讨论的频谱均为对数化和中心化后的频谱,以下同)上准确地找到半色调尖峰,从而针对尖峰进行处理,能够使处理后图像获得很好的质量,同时又尽可能地保存图像的信息。该算法对摩尔网纹,特别是低频摩尔网纹去除效果显著,这也是该算法优于其他一般去网纹算法的地方。
Presents a novel adaptive descreening algorithm for scanned halftone images that printed with AM screening. The proposed algorithm based on the rule of CMYK printing and how the moir6 patterns formed, which can exactly find out the halftone peaks in the Fourier spectrum and have them processed. Because of only having the halftone peaks processed, the restored images not only have good quality, but also greatly preserve the image information. More importantly, the algorithm is more preponderant than other current descreening methods to remove moir6 patterns （especially low frequency moir~ patterns）.
