Reducing False Triggers In Surveillance Systems Using Sensor Fusion

Abstract: Sensor fusion has been widely adopted in the last couple of years, especially in the automobile industry. Their main goal is to gain a more robust system and increase security by e.g. predicting and preventing collisions. Surveillance systems, based on video motion detection, face similar issues by having numerous problems with false triggers, particularly when there are big variations in the lighting of the scene, e.g. shadows or light beams. To address this issue, the effect of adding a radar sensor, whilst the video system is used as a black box, is investigated. There exists a presentiment that the amount of detections that are identified should not decrease noteworthy, as the two different systems complement each other. The validation is not necessarily identical with reality, however it is a clear indication that sensor fusion is more reliable than using only video motion detection.