Abstract: In recent years, with the rapid development of the economy, the indoor intelligent monitoring is getting more and more attention. This paper proposes an indoor intelligent monitoring system that fuses the audio and video due to the current states. Firstly, the system conducts pretreatments on the signals acquired by the microphone array. Then, after locating the source position, it realizes the real-time steering of the dome. Finally, the detections are made on the circumstances of the monitoring scenes using the anomaly monitoring detection technique so as to determine whether trigger the alarm. Test results in a real indoor environment indicate that the design has met prospective requirements with the sound location accuracy rate and the anomaly detection recognition rate can reach over 87.5% and 71.67%.

Key words: indoor intelligent monitoring system, microphone array, sound localization, anomaly detection

摘要： 近年来，随着经济的迅猛发展，室内智能监控系统越来越受到人们的关注。针对当今智能监控的现状，提出了一种融合音频和视频的室内智能监控系统，系统首先对麦克风阵列采集的信号进行预处理和声源定位，在估计出声源位置后，实现球机的实时转向，最后采用异常检测技术对监控场景情况进行检测，从而判断是否报警。真实室内环境下的测试结果表明，声源定位准确率和异常检测识别率分别为87.5%和71.67%，达到了预期的设计要求。

关键词: 室内智能监控系统, 麦克风阵列, 声源定位, 异常检测