Videointellect: System of video analytics

2018

The module "The Detector of the Left Objects" is certified by FSB of Russia

The Skolkovo company "Videointellect", the developer of the systems of situational video analysts, received the certificate of conformity FSB Russia on a software module "The detector of the left objects", reported in Skolkovo Foundation September 4, 2018.

During tests the detector Videointellect confirmed officially the utilization properties and compliance to all certified requirements for application in the field of transport security.

During certification tests the conditions, closest to reality — a dense passenger traffic, low contrast of objects (draining of a subject with a background) and others were modelled — Pavel Sazhin, the CEO of Videointellect company told.

Key feature of products of the company is the high efficiency of work of algorithms of video analytics in the conditions of mass accumulation of people, intensive passenger traffics, conditions of insufficient lighting, noted in Skolkovo. Such modules of video analytics as "The detector of the left objects", "The detector of the movement in a forbidden band", "The detector of the movement in the prohibited direction", "The detector of atypical change in a scene", "The detector of detection of aggressive behavior and fights" already proved on a number of objects, added in fund.

Mobotix video cameras and software will integrate Videointellect in the complete solution

The Konica Minolta company announced on April 17, 2018 agreement signature about cooperation with domestic software developer in video analytics Videointellect. Within partnership of the company will present at the Russian market the hardware-software solution on video analytics based on Mobotix video cameras. Read more here.

2017

About intellectuality of the market of video analytics

On April 20, 2017 the Videointellect company announced the vision of emerging market of the videonalitiki in Russia.

Often tolerance ranges of probabilities of false operations of algorithms (false alarm ratio) from 1 to 5% of percent are specified in requirements to the systems of video analytics for providing the necessary security level. The digits formulated in such type look as 1%.

But what these digits mean in practice? Videointellect specializes in algorithm elaboration of security for places of mass accumulation of people therefore we will review an example for such a case. At a flow in one hundred thousands of people during the working day the minimum level in 1% of false operations will make 1000 people. In terms of the working day it is on average 125 people per hour or about 2 persons a minute.

At such results the control service will be completely paralyzed by tracking of false signals within the first hours of operation of a system. In this case it is necessary to consider a psychological factor. The security service specialist accepting a large number of false signals will not be already such attentive to really taking place important event.

Values of probabilities much or better are necessary for a full-fledged work of systems in such conditions several orders less. A system Videointellect analyzes scenes and reaches parameters: 0.3 - 0.6 false operations on the video camera in day that makes thousand shares of percent of false operations.

In the market systems in which functions of search in a videoarchive are well implemented gained distribution. The employee does not have need to spend time for viewing all archive, enough, for example, to enter a photo of the required person, and the video analytics will issue all video fragments with his presence. However it is analytics a post factum. Software Videointellect works in real time that allows to react instantly to the detected situations, illegal acts. It considerably raises the security level of objects and helps to predovratit negative effects. Pavel Sazhin, CEO of Videointellect company

The product of Videointellect company creates the market of video analytics of real time. There was an opportunity to use the smart systems of video analytics for detection of situations in the most difficult conditions, places of mass accumulation of people and difficult scenes. Svetlana Chertkova, founder and founder of Videointellect company

Analysis of a psychological profile of the buyer

On March 21, 2017 the Videointellect company announced release of the version of a system of video analytics and identification of the personality for offline retail.

In cases when consultation of the manager of shop can not render assistance or be perceived as persuasive behavior, business can be helped by the systems of video analytics.

Screenshot of the software window, (2016)

A system will work by the principle of noncooperative face recognition, i.e. the buyer should not look in the camera, the analysis will be automatic. Implementation of such software will allow to distinguish buyers on gender and age, to conduct calculation and statistics of interest in goods and also to precisely identify regular customers. Other available analogs do not allow to work most effectively in difficult conditions. Pavel Sazhin, CEO of Videointellect company

At the heart of system development scientific approach - a training method of neural networks. The solution allows to distinguish persons even if the buyer does not look in the camera.

We know that it is necessary modern offline retail. Tasks which are set by owners of shops are feasible using the smart video analytical systems. We know how to resolve the existing issues - for this purpose there is a system Videointellect. Our company continues to develop a complex of solutions for offline retail, and we will offer technology for face recognition soon. Svetlana Chertkova, founder and founder of Videointellect company

Videointellect for retail

The team of scientists from Videointellect company develops the software for recognition of behavior of buyers in offline shops. A system has no analogs in the Russian and world market of video analytics thanks to a combination of the bases of science and new methods of neural deep learning. She allows to distinguish readiness of the buyer for purchase and his interest in goods and also in advance to secure itself against the preparing theft.

For March 10, 2017 Videointellect for retail is the complete solution of video analytics for offline retail.

On March 10, 2017 the Videointellect company announced release of a complex of the innovative video analytical developments "Videointellect for retail" for growth of conversion and automation of objects offline retail.

A system allows to define readiness of clients for purchase, to carry out calculation of visitors of shop and to predovratit thefts.

We develop an option of the analysis of fullness of commodity shelves in the shop. The detector will work by the same principle and with the same efficiency, as on transport objects - recognition of the objects merging with a background, the small size and when overlapping objects by people. According to experts, a system works at the level of military technologies as the efficiency supported by white papers in 98% was reached earlier by nobody. In spite of the fact that the similar solution not the only thing in the market of video analytics, software Videointellect has no analogs on the efficiency. The uniqueness thanks to which such result was achieved consists in a development method - application of algorithms of nuclear physics and deep machine learning. Pavel Sazhin, CEO of Videointellect company

The automatic analysis of fullness of commodity shelves will give the chance to control a status of stands of shop. The customized system will analyze and issue percent of fullness of shelves and upon transition through the set threshold value to give signals on the smartphone of the manager or to send messages about a stand status through certain time according to the diagram. Using data it is possible to make the diagram of devastation of shelves in different days and seasons taking into account arrangement of commodity stands.

A system will also allow to estimate efficiency of a counter, having analyzed its arrangement - whether it be a far corner or the running place. Using statistics of passability of shelves the owner of shop will be able to set the objective price to suppliers for lease of a counter and location of goods in shop.

On the one hand, the subject of video analytics in itself is not unique, solutions in the market exists enough, will not surprise with them. However nobody reached similar efficiency which is shown by Videointellect thanks to developments on the basis of algorithms from nuclear physics earlier. We often communicate with large network retailers and we suggest to test our system in actual practice. Offline retailers depending on a type of the sold goods is put by different priorities before the video analytical systems. Videointellect becomes this tool for increase in level of efficiency and automation of work in shops on all needs for each direction of this market. Developments continue, and we will present some more the innovative solutions for offline retail soon. Svetlana Chertkova, founder and founder of Videointellect company

Security on transport objects

The Videointellect company — the participant of an IT cluster Skolkovo, the developer of the systems of video analytics for difficult scenes of video surveillance and also for video cameras of non-stationary placement — announced at the beginning of 2017 the solution for security on transport objects, the being 100% the Russian development.

The company applied scientific approach to creation of a video analytical system Videointellect. The detectors developed by scientists allow to detect the left objects of the small size even on the merging background and in crowd of passengers and also to recognize asocial behavior. A system quickly transmits a signal of detection and helps to prevent threat of life in advance, specified in the company.

According to developers, algorithms for any places of video surveillance and scenarios are provided in the solution. At the same time the possibility of program and hardware integration of Videointellect with the third-party systems and the equipment is provided.

"In modern conditions of one of the most important problems of operation of objects of transport infrastructure, such as stations, subway, airports, security of people is. A key component — video surveillance systems — Pavel Sazhin noted. — However technologies of many even of global manufacturers of the systems of intellectual video surveillance, the showing excellent results in sterile conditions, become insufficiently effective in real operation on the difficult objects which are characterized by intensive passenger traffics. We created the solution cardinally different from the existing systems in the market, and we can safely say that Videointellect is a product of new generation".

According to representatives of the company, according to the results of operational testings — in the subway of St. Petersburg and Kazan — development showed the necessary level of efficiency.