基于离散正交矩的图像模糊不变性研究  

Research of image blur-invariants using discrete orthogonal moments

作　　者：高月芳[1,2] 陈国华[1] 王敏琴[1] 肖宿[1] 

机构地区：[1]华南理工大学计算机科学与工程学院,广州510641 [2]华南农业大学信息学院人机交互研究中心,广州510642

出　　处：《计算机应用研究》2009年第3期1154-1156,共3页Application Research of Computers

基　　金：广东省科技计划资助项目(2007A020100006-10)

摘　　要：提出一种基于离散正交Tchebichef矩的模糊不变矩构造算法,其对中心对称点扩散函数造成的模糊图像具有不变性。该不变矩的构造与基于连续矩构造的模糊不变矩思想相似,但其在计算过程中不存在离散近似误差和坐标变换误差,故在有噪声和无噪声情况下均具有更好的模糊不变性。仿真实验验证了本算法的可行性。This paper presented a novel method to construct blur-invariant moments based on discrete orthogonal Tchebichef moments in order to deal with blur images affected by a centrosymmetric point spread function(PSF).The structure of proposed blur-invariants was very similar to the ones proposed by Flusser et c.and Wee et c.,but the paper s derivation did not involve numerical approximation of continuous integrals and coordinate space transformation.The blur-invariants have better invariance compared with the one...

关 键 词：模糊不变性 离散正交矩 对称点扩散函数 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]