Video Surveillance System with Deep Learning Technology

Hikvision has followed up its launch of their Deep Learning NVRs with a new series of IP cameras. The new “DeepinView” IP Camera Series delivers power and intelligence to boost the value of surveillance system performance across a broad range of security and management applications.

Hikvision’s deep learning algorithms bear much deeper programming compared against conventional intelligent algorithms, which only operate on the surface level. These algorithms perform feature-learning and provide astonishingly accurate and consistent video content analytics (VCA) performance. When coupled with high-speed GPU processing, Hikvision DeepinView cameras demonstrate faster computing with large amounts of data.

Critical analytics – such as false alarm filters, facial recognition, people counting and ANPR – can now be reliably implemented in existing and new surveillance systems with versatile applications, providing a sophisticated level of tracking and alarm activation during incidents or even for pre-incident alerts. These analytics also enable customers to reduce the manpower previously required for searching surveillance footage, as well as improve business management efficiency and commercial ROI.

Human Body Detection
Hikvision deep learning technology enables the DeepinView cameras to detect human bodies while filtering out insignificant objects and movements within a scene where conventional VCA systems trigger false alarms. This is particularly useful for perimeter protection, where users often spend too much time and monetary resources locating significant alarms and relevant information.

Facial Recognition
Facial recognition can be deployed in many security and management scenarios within a variety of applications to alert system operators to the presence of persons of interest. This is achieved by the facial image modelling and similarity calculation woven into the system. This technology is becoming a crucial tool against blacklisted personnel entering to casinos, as the identity of an individual can be used to alert security guards to the presence of a known offender, enabling security personnel to deny access in that casino. In this way they can potentially prevent cheating behaviours. The technology works equally well in preventing the admittance of known offenders into venues such as sports stadiums or restaurants.

People Counting
For transportation hubs, retail stores, sports venues, visitor attractions, and car parks, the gathering and analysis of accurate visitor information can assist businesses to improve their profitability and site management.

People Counting video analytics can track the number of people who enter and exit a particular area. It can record foot traffic through a retail store on a daily basis, or monitor the number of people in a venue to ensure that health & safety limits are not breached at any one time, as in a museum for example, where crowds move through on foot.

Vehicle Management
Traffic monitoring and Automatic Number Plate Recognition (ANPR) analytics can be deployed to monitor vehicular traffic movement and enhance the efficiency of traffic management strategies. ANPR can be used to identify vehicles with listed number plates and allow them access to public and private car parks automatically. Hikvision DeepinView deep learning cameras self-learn the number plate information within a scene and recognize a larger amount of number plates than conventional ANPR systems, without the often cumbersome camera positioning adjustments.  

DeepinView Traffic monitoring analytics, when applied in populous areas and on highways, capture traffic violations such as running red lights, wrong-way driving, illegal parking, and illegal U-turns, contributing to decreased traffic congestion and placing the public and vehicle passengers at lower risk of accidents.

“Offering security professionals much deeper system functionality, the new range of Hikvision DeepinView IP cameras combine video data, immense processing power, and Hikvision’s innovative set of intelligent technologies to provide a whole new level of video surveillance performance,” says Keen Yao, VP at Hikvision International Business Centre. “With the DeepinView range, deep learning video analytics will transform standard CCTV systems into intelligent and highly-effective, HD-quality, automated detection and alert systems, to support operators and to deliver more efficient surveillance systems management.”