EN
PL
CZ
DE
IE
HU
PT
RO
SK
ES
No. 3/2021 (Jan. 18, 2021)
Even more intelligent cameras.
Modern artificial intelligence (AI) systems combine cameras (sensors) and computing devices, such as graphics processing units (GPUs). So configured AI systems analyze the image only after visual information is registered and sent from sensors to processors. However, many of the details captured by cameras, such as leaves in the background of a passing car, are not important. They pack systems with unnecessary data while consuming energy and time needed to process information. For intelligent monitoring systems to perceive the world efficiently, a change in the approach is required. Inspiration can be taken, for example, from the way in which human eyes and the brain work together so that a person can see and analyze things they consider important.
SCAMP-5d hardware architecture including a 256 x 256 PPA matrix of pixel processors, each with a light sensor, local memory records and other functional elements
During the study, researchers from Bristol and Manchester presented a convolutional neural network (CNN) in the SCAMP-5D computer vision system, and demonstrated how perception and learning can be combined to create new cameras for AI systems. The neural network was able to classify hand gestures at 8,200 fps. The CNN convolutional neural network (a type of artificial intelligence algorithm) developed by the team is able to recognize images without registering or forwarding them to processing systems.
The work was possible thanks to the SCAMP architecture developed by Piotr Dudek, a professor at the University of Manchester, and his team. SCAMP is a processor-camera system that the team describes as the Pixel Processor Array (PPA). The PPA has a processor built into each pixel, so they can communicate with each other to process data in parallel. It is an ideal solution for CNN and vision algorithms. Such cameras create not only new machine learning opportunities, but are also characterised by high speed and quick configuration. They are ideal for fast, highly agile moving platforms that can learn literally on the fly.
What tool to use to configure the 2nd generation Hikvision IP/2-Wire video intercom system?
To configure such video intercom systems, additional software iVMS-4200 from v3 is needed (the devices, except for 2nd generation Villa door stations, do not have a built-in web server). After connecting a door station and monitors, activate them using the SADP or iVMS-4200 application, assigning the administrator password, just as in the case of IP cameras or DVRs. After activating the devices and setting the appropriate IP addresses, add them to the iVMS-4200 application. Only after correct addition will it be possible to go to the "Remote settings" tab of each device and to perform its full configuration.
View of the iVMS-4200 (v3.4.0.10) window after adding the door station and two indoor stations Configuration of 2nd generation systems is possible from version v3.
Types of fiber optic splicers – what to choose?
When deciding on a fiber optic splicer, it is absolutely necessary to take a closer look at the tasks it will perform. This, among others, applies to the average number of splices performed in a monthly or annual cycle, the number of splices during one installation, and finally the type of systems to be worked on (inside buildings, metropolitan, closed LAN networks, CCTV, etc.).Depending on the answers to the above questions, the customer can choose a more or less advanced model of a fusion splicer. One of the most important processes carried out by the splicer is the positioning of the fibers before the welding process. According to this criterion, splicers can be divided into three groups:
- Fiber sheath positioning devices – the most economical option. The splicer aligns the fiber sheats, the diameter of which according to the standard is 125 μm. As a rule, this method of positioning results in a splice with slightly higher attenuation compared to other solutions, as in the case of poor quality fibers, in which the core is not centered in relation to the sheath, an ideal sheath alignment does not guarantee a good core positioning. In practice, however, the fibers sold today are of high quality and in the vast majority of cases, the quality of connections made by sheath alignment do not differ from the results of other methods.
- Devices aligning based on the geometric core – this method is used by most splicers and by many manufacturers and dealers, and is wrongly considered as positioning according to the real core. The geometric core is determined by passing visible light through the fibres. The deeper into the fiber, the lower the light intensity will be. The splicer determines the fiber core using the image analysis, with a given sensitivity. This method will obviously work for quality fibers, only then the geometric core will be the same as the real one. For fibers with eccentrically positioned real core, the splicing performance may be inferior to the sheath-based splicing solution.
- Positioning devices based on the real core – the most professional and the most expensive solution. Splicers using this method can automatically recognize the type of fibers to be joined and align them in relation to their actual core. Such splices are characterized by the lowest attenuation. Devices based on this method are several times more expensive than those using other solutions.
Signal Fire AI-9 L5875 and AI-8C L5870 splicers support two splicing methods: based on the sheath and the geometric core alignment. It can be successfully used in all types of fiber optic systems. At the moment, due to the growing technological progress and a limited number of global fiber producers, the differences between splicers with the geometric and real core alignment are being blurred. Fibers with differently positioned cores are extremely rare, and if they do occur, they do not have a significant effect on the joint alone.
Can "ordinary" hard drives be used in a CCTV monitoring system?
A hard drive is one of the key elements of the CCTV system. Its performance is critical for the reliability and quality of camera image recording. Hard drives used for monitoring purposes operate in different environments than those used in PCs. Data is stored here continuously and accounts for over 90% of the disk load. The need to simultaneously handle multiple data streams generated by individual cameras is also a challenge. Not only are desktop drives less reliable in video surveillance environments, but they do not provide the performance needed for multiple cameras either.PC hard drives:
- are designed to read and write small data blocks and handle random transfers of up to 60 TB/year,
- have a working time limited to 8 hours per day, 5 days a week,
- are not designed to handle multiple cameras,
- don't have rotational motion sensors or the RAID function, which are required in multiple-disk systems,
- don't have power management functions and cannot work in high heat conditions.
In contrast to PC drives, Seagate's SkyHawk HDDs are optimized to work with CCTV systems. They:
- designed to work 24/7/365,
- handling 550 TB data load/year,
- 90% of the time is spent on data writing and the remaining 10% on data playback and control
- saving images from multiple cameras without losing frames,
- they ensure reliability, resistance to vibrations and high performance in multi-drive systems,
- they are equipped with advanced power management functions,
- integrated SkyHawk Health Management software actively protects the system storage using prevention, intervention and data recovery algorithms.
External LTE modem.
The MikroTik LHG 4G kit RBLHGR & R11e- 4G 4G N24118 is a complete CPE device Customer Premises Equipment) equipped with an LTE modem. It is designed for an individual customers for the reception of LTE signal in the 450, 800, 1800, 2500, 2600, 3500, 3700 MHz band. It is equipped with an antenna with 21 dBi gain. Configuration is done via a web browser (basic knowledge of IP networks is required). It offers an alternative to the standard kit consisting of a modem (installed inside the building) and external antennas connected to it.The device is based on the MikrotikOS firmware level 3. It has one Ethernet port, LTE category 4 modem (download speed up to 150 Mbps and 50 Mbps upload).
Access point MikroTik LHG 4G kit RBLHGR&R11e-4G, 4G 150Mbps, 1x RJ45 100Mbps, 1x SIM N24118
Installation of dSCR/Unicable multiswitches in a classic TV/SAT systems.
Currently, there are lots of satellite receiver models available. Almost all of them support the SCR/Unicable system compliant with EN50494 and/or EN50607 using single-cable technology. The greatest advantage of this solution is the possibility to build branched systems, as in the case of terrestrial television. No star topology is required here. This means that the terrestrial television system can be easily adapted to the reception and distribution of satellite signals (taking into account the limitation of the number of users connected to one multiswitch).The diagram above shows a hybrid multiswitch system. A standard TV/SAT system with a new type dSCR/Unicable multiswitches SRQ-540 TERRA with AGC R80540 and SRM-543 TERRA with AGC R80543 to enable splitting the TV/SAT signal with a splitter. Typical multiswitches have 4, 8, 12, 16, 24 and 32 outputs. In this case, the signal at the output from a multiswitch cannot be split with an ordinary satellite splitter (operating band: 5-2400 MHz) due to two-way communication between devices (receiver-multiswitch). The multi-room service requires also separate cables to be connected to each decoder. A solution to this problem is to use the dSCR/Unicable multiswitches.
 | DS-KABV6113-RS/Surface rain shield for the Hikvision door station G74379 protects the station against weather conditions, such as rain or snow. | |
 | DS-KV6113-WPE1 IP Villa 2 gen. door/gate station G73624. Elegant design and small dimensions allow you to mount the station on a narrow post. Built-in color camera with a resolution of 2 MP with wide viewing angles (129° (H)/75° (V)) and an IR illuminator with the range of 3 m, ensures proper area observation round the clock. | |
 | The Hikvision DS-2CE11D0T-PIRL M75401 camera can be used in video surveillance systems based on HD-TVI of DVRs. It generates high quality image with 1080p resolution. Thanks to sensitivity of the image sensor 0.01 lx and built-in IR illuminator with range up to 20 m in EXIR 2.0 technology, the camera provides high quality video coverage even in low lighting conditions. | |
Worth reading
Hikvision IP DVR for large systems. DS-9664NI-I8 K22364 is a modern 64-channel Hikvision IP recorder dedicated for building large and demanding systems. With 8 SATA ports it can support up to 8 hard drives with a total capacity of up to 80 TB. It is equipped with two HDMI and two VGA connectors so that the device can be directly connected to a computer monitor or a TV set... >>>more
The K22364 recorder used in 2019 to build a large surveillance system above the building
and yard of the driver examination center
and yard of the driver examination center
RACK cabinets - full range of products
