Nowadays hackers are using many novel techniques to hide malicious files inside an image and later sent these images to target users. These malicious image files allow the hackers to get access to the system and control the victim system from their remote location. In order to prevent this, a Django interface is proposed in this research work to verify whether the file is malicious or not? The interface allows the user to choose an image after downloading from an email or any other sources. The image is then sent to the back end server to verify if there is any malicious file hidden inside the image or not? If there is any hidden file then it is extracted from the image using steganalysis, which is a technique used to extract a hidden file. This classification is done by using machine learning technique where the machine is trained using the dataset containing features of malicious and non-malicious files. Random Forest algorithm is used for classification purpose to classify the file is malicious or not. So when an executable file is hidden inside an image, the file is extracted and the features of that executable file are then sent to the already trained machine learning model. If it is malicious then a message will appear displaying that the file is malicious and if not, a message will appear, displaying that the file is safe to use. This research work is implemented and is then deployed in IIS web server in windows server 2019, so that any user can access the website and can upload single image or multiple images. The website can be accessed through mobile phones or laptops.

Malicious image stenography Machine learning