Malware Detection in Cloud Computing using an Image Visualization Technique

Malware creators generate new malicious software samples by making minor changes in previously generated samples. Here, similar features of the samples from the same malware family may be used in the detection of newly generated malicious software. In this paper, a new malware detection model based on the similarity of images is proposed. Malware detection is ensured by identifying the changes made in images. For the detection of the changes in malware images, a probabilistic framework is proposed. As a result of the experiments carried out on the two images of the program code from the same class, the proposed method accurately determines the changes made to these codes. The proposed model is tested on the Malimg dataset, and the model recognizes the changes in the software code with high accuracy.