智能空间中基于场景和本体的用户异常活动识别

Human Abnormal Activity Recognition Using Scenario and Ontology in Intelligent Space

大多数基于传感器和本体的异常活动识别研究仅考虑动作本身是否具有危险性或异常性,认为任何偏离日常活动的模型都是异常,针对可能导致的漏判或误判,提出考虑场景语义,同时加入用户本身情况因素;采集日常生活活动数据,建立相应的本体库,另外加入4种已确定的异常活动,对异常活动进行更丰富的识别;经过实验验证,异常活动识别准确率有明显提高。

Most of the researches based on sensors and ontology for abnormal activity recognition only considers whether the action itself is dangerous or abnormal,showing that the routine activities are modeled and any deviation from the activity model is recognized as abnormality,so emphasizing the scenario semantics factor and user＇ s own situation for handling misjudgement or omissive judgement.Collecting activity data of daily living and then establishing the corresponding ontology base,meanwhile adding four kinds of identified abnormal activities,it has a better recognition of abnormal activities.The result of experiment validates that the recognition accuracy of abnormal activity is improved obviously.