Robust anti-screen capture security system based on the OS-level control and the AI techniques

Abstract

Due to COVID-19 and the fourth industrial revolution, telecommuting from home or from places other than the office is spreading around the world. These cases generate new security vulnerabilities to the business sector by delivering an organization’s internal data to civil areas. Therefore, organizations now need to provide a secure working environment for telecommuters in the contactless era. Organizations and companies are starting to offer streaming data services, such as virtual desktop infrastructure (VDI) or cloud services, to provide a secure working environment for telecommuters. These services provide data without the need to download it to a third terminal. However, attacks such as screen capturing can occur when data are streamed. Therefore, the importance of a screen capture prevention technique is increasing. In our work, we present complete anti-screen capturing techniques for PCs and smartphones. The main advantage of our techniques is that we can protect the capturing of all capture tools running on Windows OS, Mac OS, Android OS, and iOS, not just a specific capture program. In addition, we propose an anti-screen capture system based on computer vision and deep learning (DL) algorithm to prevent capturing the PC’s screen data using a smartphone. The proposed method analyzes the stream from the webcam frame by frame to detect if there is a smartphone using an object detector algorithm. Whenever a smartphone is detected in a frame, it blocks the computer screen to prevent screen capturing. To the best of our knowledge, this technique is the first to apply object detection and tracking algorithms to prevent the capturing of the screen’s data using smartphones.