医学内窥镜图像横条纹噪声分析及解决方法  

Analysis of Horizontal Stripe Noise and Solving Methods in Medical Electronic Endoscope

作　　者：王威[1] 杨静[1] 刘西振[1] 

机构地区：[1]上海海事大学信息工程学院,上海200135

出　　处：《计算机仿真》2012年第1期234-237,373,共5页Computer Simulation

基　　金：国家自然科学基金资助项目(60802027);上海海事大学校基金资助项目(20090124)

摘　　要：针对CCD型医学电子内窥镜在饱和信号附近出现的横条纹噪声,图像信息的正确性影响,传统的解决方法会产生图像细节模糊和边缘锯齿化。为提高视觉质量和较低的算法复杂度、运算量,为此,提出了一种根据最近邻边缘检测信号补偿去除横条纹噪声的算法。首先从理论上分析了由于分光特性而产生横条纹噪声的原因,通过利用分光特性信息,判别像素空间信号饱和状态,并采用最近邻边缘检测,有效解决图像的横条纹噪声及边缘模糊等问题。实验证明,方法有效地去除了横条纹噪声并很好地保留了图像的边缘和细节信息,同时运算复杂度低,具有实时性处理的特性,对于实际工程应用有参考价值。For the horizontal stripe noise near the saturation signal of CCD-based medical electronic endoscope requirs a better visual quality as well as lower algorithm complexity and computation, the traditional solutions produce blurring details and jagged edges of images, therefore, the algorithm of edge detection based on the nearest neighbor remove horizontal stripes noise signal compensation was researched. The method in the paper firstly theoretically analyzed the spectral characteristics of the noise arising from the reasons for horizontal stripes, and through the use of spectral characteristic information, distinguished the signal saturation of pixel space as well as edge detection based on the nearest neighbor, which can effectively resolve the horizontal stripe noise and edge blurring, etc. The experi- ment proves that the method can effectively remove the horizontal stripe noise and reserve image edge details, and the low computational complexity makes real-time processing available, which provides great reference value for practical engineering applications.

关 键 词：医学电子内窥镜 横条纹噪声 分光特性 边缘检测 噪声去除 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]