Detection of Sensitive Textual Information in User Photo Albums on Mobile Devices
Evgeny V. Myasnikov, A. Savchenko · 2019 International Multi-Conference on Engineering, Computer and Information Sciences (SIBIRCON) · 2019
The need to identify and delete sensitive textual information in images arises in many areas. In this paper, we propose a two-step method in order to detect sensitive textual information in images or video. This method consists of text detection and recognition in an input image with the subsequent identification of sensitive words. We propose to include lexical information in the final decision to provide more stable performance in unconstrained environments. The method is implemented in a mobile application. An experimental study of several text recognition libraries is provided for well-known datasets (ICDAR 2003, KAIST, SVT, CBDAR 2011, DIQA and MIDV-500).