Camera anomaly detection based on morphological analysis and deep learning

Recently, camera anomaly detection has attracted increasing interest in order to generate real-time alerts of camera malfunction for video surveillance systems. The existing camera anomaly detection methods still haven't enough ability to detect comprehensive types of anomaly, and lack the self-improvement ability in the case of miscarriage of justice by self-learning. So, this paper proposes a morphological analysis and deep learning based camera anomaly detection method to detect comprehensive types of anomaly. Morphological analysis is used to detect simple camera anomalies to accelerate the processing speed, and deep learning is utilized to detect complicated camera anomalies to improve the accuracy. The experimental results show that the detection accuracy of the proposed method achieves more than 95%.