医学图像分割及恶意样本分类研究

摘要

目前因为恶意代码生成手段处于连续更新阶段，持续生成各种能够越过传统检测技术屏障的新式恶意代码，这使得反恶意代码的任务变得非常艰巨。因此，对恶意代码进行快速、准确的分类成为了反恶意代码的重要手段之一，可以为对恶意代码的追踪、掌控与删除奠定良好的技术基础。所以，如何有效进行恶意代码分类工作成为社会重点研究问题之一，本文提出了一种新的恶意代码分类方式，并建立了恶意代码分类模型，实现了比较好的分类。

医学图像分割对于医疗诊断和病理学研究具有重要意义。为了快速有效的处理医学图像，本文提出了一种基于Spark分布式处理医学图像的方法。

本文主要做了一下工作：

1.恶意代码部分：本文建立了一种新的恶意代码分类的模型，通过提取PE文件中函数字节码作为特征，并利用最大公共子序列和极大团算法建立分类模型，取得了较好的分类效果。

2.医学图像分割部分：在本文中，首先建立了一种基于改进的模糊核聚类和CV模型的医学图像分割模型，然后用Spark对医学图像进行了批量处理，最后进行了模拟实验，其结果表明本文提出的方法实现了高效精确的分割。

 

At present, because malicious code generation is in a continuous update stage, it continues to generate a variety of new malicious code that can cross the traditional detection technology barrier, This makes the task of anti malware very arduous. Therefore, the rapid and accurate classification of malicious code has become one of the most important means of anti malware. It can provide a good technical basis for tracing, controlling and deleting malicious code. Therefore, how to effectively carry out the classification of malicious code has become one of the key issues in the society. A new classification method of malicious code is proposed in this paper, and a malicious code classification model is established to achieve better classification.

Medical image segmentation is of great significance to the research of medical diagnosis and pathology. In order to deal with medical images quickly and effectively, a method of distributed processing medical images based on Spark is proposed in this paper.

The main work of this paper is:

1. malicious code part: This paper establishes a new classification model of malicious code. By extracting function bytecode in PE file as a feature, and using the largest common subsequence and maximal clique algorithm, we establish a classification model, which achieves a good classification effect.
2. part of medical image segmentation: in this paper, firstly, a model of medical image segmentation and fuzzy kernel clustering algorithm based on improved CV model is established, and then use the Spark of medical image for batch processing, finally through simulation experiments, the results show that the proposed method achieves efficient and accurate segmentation.

AI创作