基于小波粗糙集算法的图像拼接研究  

Image mosaic research based on wavelet and rough set algorithm

作　　者：李宏升[1] 张健[1] 

机构地区：[1]黄淮学院信息工程学院,河南驻马店463000

出　　处：《液晶与显示》2014年第2期298-303,共6页Chinese Journal of Liquid Crystals and Displays

基　　金：河南省科技厅项目(No.122102310557)

摘　　要：为了提高图像拼接的质量,采用小波粗糙集算法。首先,基于自适应小波变换对图像分解,通过小波方向的对比度设定阈值来平衡图像内各个对象的差异。接着,分解图像特征点的粗糙决策属性来划分重叠区域,通过空间的划分,把上近似在属性子集的划分区域划分到对象集合中,下近似有可能划分到其他对象集合中,比较计算参考图像与输入图像特征向量因数集的欧氏距离约简规则选择特征点,并把输入图像中欧氏距离最小的点作为参考图像中特征点的匹配点。然后,通过平均值法消除拼接缝隙,实现图像的上下、左右以及斜方向的无缝拼接。最后,给出了算法流程和评价指标。实验仿真表明,本文算法拼接缝隙区域过渡光滑,视觉效果连续,没有出现重叠鬼影,本文算法的信息熵为33.872 1bit,处理时间为2.436 8s。满足拼接结果中对执行时间少、信息含量大等要求。In order to improve the quality of image mosaic,a wavelet and rough set algorithm is estab- lished. First, image is decomposed by adaptive wavelet transform, directions of the wavelet of image are set threshold to contrast and balance differences in individual object. Second, feature point is divi- ded the overlapped area, approximation properties in a subset of the divided region are divided into collection of object, lower approximation could be divided into other objects in the collection, reference and input feature vector factors of image are compared and ruled to Simple Euclidean Distance. Third, gap of image mosaic is eliminated by average method and disappeared in other directions. Finally, the whole process is described. The experiments show that this algorithm can made gap smoothly, visual appealing, without overlapping,and entropy of information is 33. 872 1 bit, time are 2. 436 8 s. It takes less time and has large information.

关 键 词：小波变换 粗糙集 特征点 拼接 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]