Most of the companies produce or make their safeguard systems produced based on security concept preceded by risk analysis.
The objective is to develop a profitable and adequate level at the same time.
Thus they put their security requirements, their assumptions on such systems and the methods of implementation in a comprehensive frame. But what effect can multi-megapixel cameras take on this concept? What new objectives can be formulated and what new proceedings can be applied to the implementation of these objectives?
Growing expectations
The series of crime of recent period have once again placed attention on video surveillance systems and the increasing expectations on them. More and more we can hear about opinions criticizing the quality of surveillance recordings.
Why?
Influence of the Consumer Electronics Devices
The reason for this is that today we have got used to pictures and recordings. For as a result of the speedy development of digital imaging consumer electronics devices can present considerable improvement in quality from year to year. Recently one can buy mobile phone devices with 10 megapixel camera, and the assortments comprise even Full-HD TV-sets or 10 to 20 megapixel digital cameras.
We're getting accustomed to having the quality of pictures, recordings and TV broadcast are far much better than it was a year or some years ago.
However this has not attracted any considerable change in the quality of surveillance recordings during recent periods, for analog systems have reached their limit in video quality.
Technological Improvement
Nevertheless the technology is ready to infiltrate into safeguard technique, showing up video surveillance devices and equipments suitable to the expectations of the time.
For today’s image perceiving devices exceed considerably the parameters of analog cameras. We have megapixel optics available, fully developed, scalable image compression procedures, as well as auxiliary storage of adequate size.
And technological development is accompanied by substantially falling prices.
Tendency of prices
The consequence is that the rate of price decrease of both megapixel or multi-megapixel IP cameras surpass that of analog cameras.
In the near future this tendency will result the appearance and blasting spread of multi-megapixel systems available at the price range of analog cameras.
But what effect do multi-megapixel cameras have on the protective concept of video surveillance systems?
For answering this question we have to flash back to the traditional designing model applied for analog cameras, the so-called designing based on experience.
Designing Based on Experience
Let’s see how a video surveillance system is made generally.
Customers often tend to define the areas to be covered, the potential settlement location; beside they raise expectations on recognizability.
The designer prepares the device allocation plan according to the learned expectations and the field research, indicating the settlement heights and the aperture of the cameras, choosing even the appropriate object lenses.
But what does the system serve for?
An analog system seldom fulfills expectations for full person identification. We may count with two possible ways for developing the system that meets all expectations. One way is to mount extra cameras on required locations, exceeding the number of cameras generally accepted and applied. Nevertheless, this may come up against both economical and aesthetical difficulties.
The other way is to use cameras of higher resolution. However to make the best of the possibilities of megapixel or multi-megapixel cameras, user expectations may be defined precisely or more exactly.
Objective of Video Surveillance
As a first step we have to define what the objective of the video surveillance is.
It definitely is to supply the recording of the inspected object in the adequate quality.
Levels of Recognizability
The object of the surveillance may be ranked into an extremely wide range of possibilities. It may be bank note, faces, registration plates, etc.
Depending on the object of the surveillance and the expectations, we may adjust different settings for each camera and settlement location.
Levels of Recognizability – Horizontal Resolution
The least to expect is to be able to only detect the object of surveillance. Such an application may be the function of the surveillance of assembly of people.
The next level is to expect limited recognizability. This means to be able to recognize or to identify the object of surveillance. But provided that the circle of the objects that appear on the recording is known. Such adaptation stands for in-house office cameras screening inner abuse.
The third level is for genuine recognition and identification. The object of surveillance is recognizable and identifiable based on the recordings. Such application is for the surveillance of disbursement locations.
The characteristics of the different levels are the number of pixels required for fulfilling the level of recognizability depending on the object of surveillance that is the horizontal resolution of the object (Po). This quantity of pixels is the result of the conciliation of the user and the designer. This is independent of the system to be used.
From this one can easily define the width of the field of vision (W) using a simple proportion, which makes the quantity of pixels belonging to the object of surveillance and the level of recognizability insurable.
Provided the width of the target (O) and the horizontal resolution of the applied camera (Pw), we have the result.
But what does the width of the field of vision serve for?
The effect of modifying the area of surveillance
Using a camera of a given resolution the quantity of pixels is constant, a value independent of the angular aperture. By modifying the angular aperture, the quantity of pixels does not change, however the dimension of the area per pixel increases. Since this decreases the quantity of pixels for the identification of the target, thus the increase of the optic angle decreases the distance of recognizability.
Effect of the angular aperture, designable recognizability
But the width of the field of vision (W) that can assure expected horizontal resolution is constant. This leads to that given the angular aperture and the width of the field of vision, the area providing the expected level of recognizability is designable.
Designing Based on Recognizability
A system corresponding to the expectations of the client can be solved, such as the information on the suitability of the video surveillance system right before the realization, in the stage of designing.
The focus of the designing is inevitably shifted towards an implementation based on editing and computing - likewise the case of fire-alarm systems.
The width of the field of vision pertaining to the level of recognizability can be computed for the cameras of different resolutions. The necessary amount of cameras and the optimal angular aperture can be defined by the use of this datum and the areas to be covered. By using cameras of 1.3 MP, we can achieve considerably better results than by analog systems. However by using the generally accepted amount of cameras we still cannot fulfill all the expectations of face recognition.
But by further increasing the resolution we may formulate new targets as well (e.g. bank note identification or full coverage). The width of the field of vision of a camera of 10 megapixels already allows face and bank note recognition that fulfills user expectations and we may even decrease the amount of cameras.
| Designing Based on Recognizability: 1.3 megapixels | Designing Based on Recognizability: 10 megapixels |
 |
|
The designing may be influenced by many factors to be considered by the designer, such as luminance, features of the applied optics, or depth of focus.
As a summary we can formulate that multi-megapixel cameras may alter the security concept by their considerably higher resolution. On one hand they provide an opportunity for the implementation of bank note recognition or full face recognition coverage. On the other hand, staying with the original expectations, a video surveillance system can be built using even less devices. The security concept may be completed by new expectations or design methods that open the door to quantifiable quality expectations, thus exploiting the possibilities involved in the designing based on recognizability.
