基于NSCT和PCNN的医学图像伪彩色处理技术研究  

Research on Medical Image Pseudo Color Processing Technology Based on NSCT and PCNN

作　　者：李江 LI Jiang(School of Computer,Beijing University of Technology,Beijing 100124)

机构地区：[1]北京工业大学计算机学院,北京100124

出　　处：《计算机与数字工程》2021年第1期169-174,共6页Computer & Digital Engineering

摘　　要：由于人眼对灰度的识别能力远远低于对彩色的识别能力,故在症状诊断及治疗时对灰度图像进行伪彩色处理,医生可以获得更多、更准确的信息。结合非下采样contourlet变换(NSCT)的平移不变性、多尺度、多方向特性和脉冲耦合神经网络(PCNN)的同步脉冲发放、捕获特性,论文提出一种基于人类视觉特性的医学图像伪彩色融合方法。首先,根据源图像的灰度信息,计算源图像公共区域。然后,使用NSCT和PCNN处理源图像得到灰度融合图像和融合图像来源信息。最后,根据人类视觉特性,使用源图像公共区域和融合图像来源信息,对灰度融合图像进行伪彩色处理得到伪彩色融合图像。实验结果表明,论文算法得到的伪彩色融合图像特征明显,有利于人眼的识别,与灰度融合图像相比,突出了融合图像的公共区域和来源信息,有利于发现病灶。Since the human eye's ability to recognize grayscale is much lower than that of color recognition,pseudo-color pro⁃cessing of grayscale images during symptom diagnosis and treatment allows doctors to obtain more and more accurate information.Combining the non-subsampled contourlet transform(NSCT)with translation invariance,multi-scale,multi-directional character⁃istics,and pulse-synchronization and acquisition characteristics of pulse coupled neural networks(PCNN).This paper proposes a medical image pseudo color fusion method based on human visual characteristics.First,the source image common area is calculated based on the gray scale information of the source image.Then,the source image is processed using NSCT and PCNN to obtain a gray⁃scale fused image and fused image source information.Finally,according to the human visual characteristics,the pseudo-color pro⁃cessing is performed on the gray-scale fused image using the common image of the source image and the fused image source informa⁃tion to obtain a pseudo-color fused image.The experimental results show that the pseudo-color fusion image obtained by the algo⁃rithm has obvious features,which is beneficial to human eye recognition.Compared with the gray-scale fused image,the common area and source information of the fused image are highlighted,which is helpful for finding the lesion.

关 键 词：医学图像融合 人类视觉特性 NSCT PCNN 伪彩色增强 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]