Text Data Anti Pinch Technology Based on Visual Font Filtering Binary Security Detection
Kun Li, Yuqi Li, Yong Ge · 2024
Text entrainment attacks target users by embedding malicious information in text data. Due to malicious information being hidden in normal text data and difficult to detect, this attack method has a high successful rate. This paper proposes a new anti pinch technique for binary text compression attacks. Firstly, one-dimensional ordinary text data is transformed into two-dimensional data based on the dot matrix or vector graphs. Second, the text character's library based on FreeType is designed, which utilizes the characteristics of FreeType's character library to filter glyph and vector data during font file parsing. Finally, specific binary format characters or vector characters are checked in these data to quickly and efficiently discover potential embedded binary data, effectively preventing malicious attacks by binary code during transmission, storage, and use, ensuring the security and integrity of text data.