隨著人工智慧技術在安防行業的滲透和深層次應用，AI+安防已成為一種新趨勢。按照應用場景來看，人工智慧在安防行業的應用主要分為卡口場景，即人臉/人員卡口和車輛卡口和非卡口場景即普通治安監控場景。
With artificial intelligence technology gradually penetrates and deeply applies in the security industry, AI+ security has become a new trend. In terms of application scenes, there can be divided into two types, namely, person or vehicle bayonet and common security monitoring non-bayonet scenes.
那麼AI在安防行業有哪幾方面的應用呢？一起來看看吧。
So what are the AI applications in the security industry? Let's have a look.
人臉身份確認
人臉身份確認應用屬於卡口場景仿人AI應用，在關鍵監控點位安裝人臉抓拍攝像機，通過後端人臉識別伺服器對抓拍到的人臉進行比對識別，確定該人員身份。目前人臉身份確認有與人臉黑名單庫比對識別和靜態人臉庫進行比對識別兩種。
人臉黑名單動態布控應用主要利用人臉抓拍攝像機從高清/超高清視頻畫面中使用深度學習模型檢測並抓拍人臉照片，然後提取人臉深度學習特徵向量與黑名單庫人臉進行比對並實現報警提示；人臉靜態比對是指使用深度學習模型檢測並抓拍人臉圖片，然後將提取的人臉深度學習特徵向量與靜態庫中的人臉比對識別，確認該人臉身份。
The face identification application belongs to the human-imitative AI bayonet application. That means installing face camera at the key monitoring point and uses the back-end face recognition server to confirm identification. At present, face blacklist matching and static face database matching are the major two types of face identification.
人臉身份驗證
人臉身份驗證屬於卡口場景應用，人臉白名單應用屬於人臉身份驗證應用。人臉白名單應用已經在很多行業落地，比如人臉門禁、人臉速通門、人臉考勤、人證核驗等，廣泛應用於企事業單位、各類園區等場景。除了實現基礎的人臉識別應用外，人臉身份驗證還使用活體檢測等手段，防止通過照片、視頻、面具等人臉假冒行為，切實保障出入口人員安全管控及日常人員管理等。
Face authentication belongs to bayonet scene application, and face whitelist is one of the representative examples. Nowadays, face whitelist can be seem in many scenes include face access control, face attendance, witness verification, etc. Except basic face recognition, Telpo TPS980 with liveness detection function can effectively prevent counterfeit behaviors like photos, videos, masks, etc. and further ensure the security control and daily management of the entrance and exit personnel.
車輛識別應用
車輛識別應用屬於卡口場景應用。車輛識別技術是應用最成熟、效果最明顯的技術之一。藉助遍布全國各地交通要道的車輛卡口，車牌識別使得「以車找人」成為現實，成功協助警方破獲各類案件。車輛識別技術已經從初級的基於車牌的車輛識別應用階段發展到車型識別、套牌車識別等精準的車輛識別應用階段。
The vehicle identification belongs to the bayonet scene application. Vehicle recognition technology can be seem as the most mature and effective technologies. With the help of the vehicle bayonet which is widely distributed across the country, license plate recognition makes finding people by car become a reality, and successfully helps the police to solve various kinds of cases. Vehicle recognition technology has developed from vehicle recognition to vehicle type recognition, license plate vehicle recognition and so on.
行為分析
行為分析可輔助安防應用。通過行為分析系統對人員的異常行為進行分析處理，可應用於重點區域防範、重要物品監視、可疑危險物品遺留等行為的機器識別，也可對人員的異常行為進行報警，極大提升了視頻監控的應用效率。還可以實現對群體的態勢分析，如人群密度分析、人員聚集分析等，對重點區域或人員聚集較多的場所態勢進行分析，防止人群事件發生，做到提前預警、及時處置。
Behavior analysis can assist security application. Personnel abnormal behavior can be analysised by the behavior analysis system and used in recognizing key area prevention, important goods monitor and abnormal behavior alarm, greatly improves the application efficiency of video surveillance. Besides, it can also realize group situation analysis, such as the analysis of crowd density, people gathering and so on, so as to prevent the crowd events and to achieve early warning and timely disposal.