恶意代码可视化RGBA图像方法研究

Research on RGBA Image Visualization Method for Malicious Code

随着网络规模的发展,关于恶意代码的变种检测、家族分类等需求的日益增多,出现了多种将恶意代码可视化为灰度图像后进行分类的方法研究,但现阶段的可视化方法使用的特征较单一并且灰度图像可容纳特征和信息过少。针对目前恶意代码可视化为灰度图像的方法存在特征不足和信息含量少等问题,提出使用RGBA模式彩色图像来将提取的多种特征进行储存和融合,对已有的可视化方法进行改进来满足恶意代码的分类需求。

With the development of network scale,there is an increasing demand for variant detection and family classification of malicious code.Therefore,various methods of visualizing malicious code into grayscale images for classification have emerged.However,the current visualization methods use relatively single features and grayscale images can accommodate too little feature and information.This article addresses the issues of insufficient features and insufficient information content in the current methods for visualizing malicious code into grayscale images.It proposes to use RGBA mode color images to store and fuse multiple extracted features,and improve existing visualization methods to meet the classification requirements of malicious code.