Abstract

One of the risks of information security of companies is the risk of information leakage through the screen photography. Often this happens using a mobile device. In view of this, it is worth the task of automated detection of this fact of violation of the security protocol. Classically, the phone search method in the frame is to consistently study the entire frame area. However, the specifics of the phone search in this task lies in the partial overlap of the brush with a brush, which makes the phone to be small among other background objects in the frame. Therefore, a combined algorithm is proposed, combining both the hand detection method, and the method of finding the phone. Issues of using hand detection methods are given in the image, as well as methods for finding rectangular areas. Finding on the image of the hand allows you to narrow the search area of the phone. For this, the boundaries of the rectangular search area, the coordinates of which depend on the coordinates of the keypoints of the hand are determined. Further phone search is made only in this area. This allows to increase the accuracy of the detection of the phone in the frame. The article provides an assessment of the results of the proposed algorithm. To do this, a comparison of the accuracy and speed of the proposed combined algorithm with existing ones. The results of the comparison are reduced to the table. The assessment showed that the use of a combined algorithm allows to improve the accuracy of the detection of the phone with a small loss of speed of processing of the entire frame.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call