Kassiopea

Kassiopea is a software package on collecting of statistical data and their analysis for public organizations, transport, the sphere of trade. A system automatically processes a video flow from cameras and analyzes video by the rules set by the user collects statistics and transfers to the central server, presenting in the form of convenient and clear diagrams that allows to reveal, in turn, quickly important data in the business analysis.
The complex uses unique technologies of tracking: algorithms of Synes are capable to isolate necessary objects from a dense flow and to accompany them for a long time. Work accuracy - not less than 95%.
The complex gives the chance to estimate the following key indicators:

• run coefficient (ER - Entry Rate);
• conversion coefficient (CR - Conversion Rate);
• effectiveness ratio of advertizing, work of personnel, etc. (KPI - Key Performance Indicator) and others.
  

The uniqueness of development consists also in a wide choice of modules which the customer can pick up depending on specifics of the sphere of the business and requirements:

• calculation of visitors;
• the analysis of behavior in queue;
• loitering detector;
• qualifier of behavior;
• attention assessment;
• analysis of behavior of visitors.

At the heart of system operation – the unique detector distinguishing on video of the head of people. Each detected person is followed in a frame, at the same time such noises as shadows, flares and inappropriate objects (goods, home decoration) are ignored. In other words, cameras look in a floor and consider people on the heads literally of this word. Installation of cameras depends on a specific objective. Monitoring of length and waiting time of queue at the checkout is necessary? Then it is necessary to hang up the camera directly before cash desk with a small stock of the overview: installation of the individual camera is necessary for each cash desk.

Recognition and calculation of people at the checkout

Apparently in this drawing, a system detected and recorded three buyers at the checkout. Many buyers saw declarations near cash desks where it is told - "if queue in cash desk is exceeded by three persons, call number and tell the administrator". Now technologies allow to automate this process. With exceeding of in advance specified number of visitors in queues, the mechanism of a system will create the visual message and will send to phone of the head or duty operator.

The red curved lines going from the heads of buyers are transits of visitors on shop. Having such opportunity, a system can construct popular routes, give clear idea of flows of buyers, at the same time defining "dead" zones of shop.

Route chart

The thermal card - one more tool of the analysis of behavior of buyers in the field of retail. The card forms by summing of time during which in a certain point there was an object. The red zone on the thermal card means the most often visited point, yellow and green – (Fig. 3) which were less visited in decreasing order. Result – selection of sections where the client is delayed the most long time.

Thermal card

Such thermal card allows to carry out the analysis of show-windows/shelves for the calculation of more marginal products, optimization of work of personnel in a trading floor, i.e. the user obtains reliable and complete statistical information about the most attractive zones in shop. Such statistics allows to prove to lessors cost by lease of a payment for the area, in particular.

The system of calculation of visitors is organized (Fig. The camera on an input) as well as monitoring of queue at the checkout: the camera it is put on an input in shop and looks in a floor. The camera considers all people and keeps statistics: quantity come into shop, but not passed by how many from them reached cash desk or department.

The camera on an input

Key advantages:

• mobile notifications;
• data encryption (HTTPS);
• differentiation of roles and user rights;
• automatic monitoring of working capacity;
• the mechanism of preserving and sending statistics in case of low-quality connection or long-term break;
• backup of all collected statistics;
• the centralized data storage on a cloud server.

On materials of Synes LLC