A Fast Recognition Algorithm for Photoelectric Peeping Equipment

Abstract

The spreading use of concealed miniature photoelectric devices threatens the information security of interior activities. Information leaking from interior environments like meeting room, drama theatre, technical lab and private space may cause serious social problems or commercial losses. To solve the critical problem, this study proposed a practical system based on the "cat-eye effect" to detect the interior optical targets in real time. The proposed method reduces false positives based on strength of the echo signals and shape of the laser cross section. A morphological criterion is presented to determine the true target. Laboratory experiments have been performed repeatedly with different kinds of interference and background of different complexity to validate the proposed method. The results indicated that the true target was recognized accurately and rapidly. Moreover, the recognition results were demonstrated not influenced by external light illumination conditions. The average runtime was 67ms which is superior to the recent methods in the literature.