Nowadays, there are several approaches to reduce server’s and client’s equipment cost. Read the article and find out how to build the best video surveillance system with Macroscop IP camera software.
Any end-user selecting a video surveillance system, as well as any installer selecting the system’s components and installing them, is seeking a balance between quality and price. System installers are taking various steps in order to reduce the cost: they choose cheap components, use free software, skip using video analysis functions for saving on server hardware, limit archive recording time for saving on hard disks, etc.
Server is one of the most important and expensive components of the video surveillance system. That is why if you manage to reduce these costs, it will result in significant savings on the whole system. One of the most effective options for savings without sacrificing the quality and reliability of a video surveillance system is limiting the required computing resources of a server and a client’s video equipment through the use of professional software.
There are several solutions for achieving economic efficiency.
1. The use of video analysis technology of compressed video streams from IP cameras without fully decoding. By using this technology, an end-user can connect larger numbers of IP cameras to a single server, or use low-budget servers.
The decompression of video streams takes place in a few stages. The last stages require the intensive use of computing resources. These very last stages are not necessary to perform if the compressed video streams analysis technology is being utilized.
For example, algorithms allow you to analyze data and identify or search an object by color characteristics without full decompression.
As a result, the processing speed is increased. At the same time, on average, the CPU load is reduced by 3/4. This also means that a particular server can process videos from multiple cameras, which results in significant savings on computing equipment.
2. Using MJPEG compression format
There are many compression formats. MJPEG, MPEG-4 and H.264 formats are the most popular in video surveillance systems.
Compressed MJPEG Video stream is a sequence of static images format - JPEG images. Compression takes place individually for every frame or snapshot. This provides complete independence from individual images. The quality of images is maintained at a good level while playing a video archive: you can get high-quality images from MJPEG for a footage of events. At the same, time this format requires fewer resources, but uses them very intensively and requires large disk spaces.
Compression is performed for each individual frame as well as for a series of frames (intra-frames) for MPEG-4 and H.264 formats. H.264 video format (optimized MPEG-4 or MPEG-4 part 10) is not a sequence of individual images, but a chain of related data - streaming video. The advantage of this format is that not all individual images are stored, but only an initial image is stored along with all of its further changes.
When a significant portion of an image does not change and remains the same, it results in a much smaller size of a final movie than in MJPEG format. If the MJPEG format can send a set of 200 Kbytes for each image, H.264 format sends one original image of 200 Kbytes, and with all its subsequent modifications, it is much smaller in size. A final video clip in H.264 format will be at least 70-90% smaller than in MJPEG format. This also means that a smaller network capacity is required for transmission. At the same time, H.264 format processing requires more computational resources.
Overall, between MJPEG format and both MPEG-4 and H.264 formats, the first requires less computing resources, but increases the load on a video system network, and requires larger disk spaces.
3. Inter-dependence of quality and type of scene in a video
Some software packages use the so-called 'flexible' approach in changing video recording quality: when there is no action on a screen, a video archive is recorded in low resolution. As soon as the action starts, the recording quality is improved. This technology allows you to unload servers and computers and to reduce the load on a network, as well as reduce the required disk space for backup storage.
4. Alternative cameras video stream processing on a server
Most software supports dual-flow function, which means that they work with both a primary and alternate stream. The end-user can set-up parameters for each stream as well as scripts for working with them. For example, a primary stream typically has a high resolution and can be used to be recorded to an archive. However, an alternative stream has a lower resolution, and may be used for an analysis on a server. Analysis of a video with lower resolution requires fewer computing resources, and thus the load on a system server is limited.
5. Utilizing mobile and web-based clients for savings
Nowadays, mobile and web-based applications working with video systems are in high demand. That is why software developers are focusing not only on their user-friendliness and the cross-platform usage of those applications, but their impact on network and server congestion as well.
It is well-known that the H.264 format is the most popular for video compression. Most mobile and web-based clients do not support this video format for video surveillance systems. A special video transcoding from H.264 to MJPEG procedure is required to be able to get a picture for users. This procedure consumes a lot of resources and results in an additional load on server equipment.
Macroscop developers created algorithms allowing web-based and mobile clients for iOS, Windows and Android-based devices to perform video stream processing in H.264 format. It leads to savings of server computing resources for mobile and web-based applications.
Usage of video card resources by software application means that data is processed in the GPU.
In order to reduce central processor load, developers suggest transferring part of operations to a video card (processing is done by a graphic processing unit (GPU)).
A significant technological breakthrough has been achieved in graphics processors development in the recent years. The use of parallel processing technology by GPUs turn them into full-capacity computing devices. A GPU is designed in such a way that most of it comprises computing units used for data processing, which use a central processor in a much smaller proportion. The continuous technological improvement of GPUs resulted in GPUs having first caught up with the CPUs’ processing power, and afterwards exceeding them. That is why an idea of partially transferring computing operations from CPUs to GPUs has occurred to programmers who are working on video surveillance software. Modern graphics cards possess decoding mechanisms hardware of significant power. Ultimately, they enable client computers to unload. Also, the use of GPU resources allowed developers to implement the technology of video streams image decoding to be done directly on video cards.
Usage of multi-stream function
The use of multi-stream software functions not only saves servers but client’s equipment computing resources as well. It is possible to configure each stream’s parameters individually selecting less resource-intensive devices for displaying purposes. Such an approach requires less computer capacity.
Nowadays, there are several approaches to reduce server’s and client’s equipment cost. All of them aim to save on computing hardware resources.
Savings on server’s hardware can be achieved if:
Savings on client’s equipment is achieved by using an alternative video stream with lower resolution and transferring some of the computational operations onto graphics cards.
Download Macroscop Demo to check how it works!