Quick search for a needle in a haystack, or how to properly organize work with a video archive.
Technologies do not stand still, their development is daily pushed by the ever-increasing demands of consumers. Indeed, if 3-4 years ago recording at a speed of 50 fields per second on 16 channels did not raise any questions, now the user wants to have Real time recording and even with a resolution of 4CIF. At the same time, the depth of the archive of continuous recording should be at least 1 month. Naturally, these requirements are often much too high and do not correspond to the status of the object, but convincing the customer of the inappropriateness of these requirements is quite problematic. There are usually two arguments: the client is always right and if the client is wrong, then see point 1.
In principle, providing the necessary amount of disk space is not the most difficult task. For example, there are video recorders with a simple connection of expansion units with two hard drives of 750 GB each, allow you to have an archive of up to 22.5 TB.
The problem will arise later. How not to get lost in this amount of information and find the desired fragment in the shortest possible time?
Let's try to figure out this issue and separate the needle from the hay.
Step 1.
It is necessary to reduce the amount of «hay».
So, the first thing to do is to talk to the client again and try to convince them of the correct formation of the video information archive. We have already said that this task is difficult, but it is worth trying in any case, because this is the only way we can get rid of gigabytes of unnecessary information.
A simple example. Let's take any object, be it a bank, a government agency or just a school. By analyzing the activity during the day (week), we can determine time intervals with high and low activity. Filling hard drives with recordings of empty corridors is an absolutely pointless exercise. At most, you can record at a very low speed, provided that the recording speed automatically increases in case of an alarm.
The alarm can be initiated by external sensors (door opening, volume sensor, etc.) or a built-in motion detector. The latter is not a special function in modern recorders, but when choosing one or another device, you should carefully study the capabilities of the motion detector in order to use it as much as possible to save disk space.
Here are some points to pay attention to:
the number of split points. The greater the number of points, the more accurate the setting will be. Thus, the image from the camera can be divided into several activity zones (for example, several doors in the corridor), and detection will occur only at the moment of activity in these zones;
motion detector sensitivity. Determined by the number of points with changes, at which the triggering occurs;
several motion detector masks. A really important function, which not all recorders have, is the presence of different detector settings at different periods of time. For example, at night and during the day it is possible to assign different detection and sensitivity zones;
filtering by response speed. An interesting function that allows you to filter out slow movements, for example, if you need to record only traffic movements, ignoring pedestrians. And also to respond to very slow movement, when an intruder tries to deceive the motion detector by moving very slowly. The principle of the motion detector is based on comparing adjacent frames. Some video recorders will also allow you to compare specified reference frames. Thus, if the device does not detect changes in adjacent frames, it will record a change in the situation after some time;
Naturally, the operation of the motion detector should not limit the operation of the device during normal recording. For example, preliminary recording (up to 1 hour) can be carried out both during normal registration and during alarm.
Thus, the use of a flexible motion detector reduces the amount of useless video information.
Step 2.
It is necessary to increase the «needle».
This postulate is also clear, it is difficult to find a needle, since it is comparable to the neighboring straws, but if you highlight the needle, it will be much easier to find it.
A solution to this problem may be to separate alarm and normal recordings into different video archives. It is necessary to use video recorders that allow duplicating only alarms to external drives using the active mirroring function. Thus, it is possible to build a system in which an archive of operational information will be written to internal hard drives, alarm records from which will be duplicated to external drives, creating an archive of only alarms (external or by motion detector) for their long-term storage.
At the same time, the alarm log allows you to quickly search for the necessary record, selecting really important records from the general archive of video information.
Step 3.
Sometimes you need to use a «magnet»
The search function in video recorders serves as such a magnet for finding the necessary fragment.
If the date and time of the incident are known, the search will not take much time. But when such information is missing, the search without special capabilities can drag on for a long time. Let's go back to the beginning and imagine that you have a 22.5 TB archive. How long will the operator view such an archive?
This process can be sped up at least twice if you simultaneously play video information from the beginning and end of the recording on two monitors. This is possible with a dual multiplexer video output.
Another option is to search by motion detector. After selecting the desired area, the video recorder will search for all movements occurring in the selected zone. This function is convenient if you know in advance where the search object was located.
If you need to provide law enforcement agencies with the original fragment of a video image, the search function on hard drives will allow you to find the media with the necessary information and extract only it.