Prospects for the development of the video surveillance industry: the capabilities of modern video analytics systems
According to the agency Transparency Market Research, by 2022 the total turnover of the video surveillance market will reach 71 billion US dollars and will increase by at least 16.5% annually. Although the demand is currently fueled by factors such as the increased risk of terrorist attacks and the continuing tightening of legal requirements for the security infrastructure and access control, in the near future the focus will shift to industries such as retail, industry, health and others. The reason for such changes is the rapid development of intelligent video analytics systems that can significantly expand the functionality of conventional IP cameras.
The world's first video surveillance system was developed by the Siemens concern in 1942 specifically for the Peenemünde test site, where Fau-2 missiles were tested. It was designed by the engineer Walter Bruch - the “father” of the PAL analog color television standard, adopted in 1966 as a state television in Germany, the United Kingdom and other Western European countries. With all the advantages, this solution had one major drawback: the operator could not leave the workplace for a second, as the signal was not recorded in any way. This continued until 1951, when video recorders (Video Tape Recorder, VTR) appeared, created by Ampex under the direction of electrical engineer Alexander Ponyatov, who emigrated to the United States in the late 1920s.
The first Ampex VR 1000B VCR
The importance of this invention for the industry is difficult to overestimate, because, according to IMS Research, even a physically healthy, well-rested person begins to miss up to 45% of critical events occurring on the monitor after 12 minutes of work, and after 10 minutes, this figure tends to a mark of 95%. Given this fact, the IP boom that occurred in 2014 does not seem so strange. Despite the emergence of AHD (Analog High Definition - High Definition Analog Video), network cameras turned out to be much more popular than analog solutions that were common at the time. And it's not just the picture quality. The main thing here is the vast opportunities that open up digital devices in front of their owners. After all, if before the analysis of the information received was carried out only after the fact, today security systems allow solving a wide range of tasks in real time. Modern IP cameras are strikingly different from their progenitor Axis Neteye 200, created back in 1996.
')
IP camera Axis Neteye 200
Many models of network cameras are already completely self-sufficient, and the IoT concept provides the ability to create a combined infrastructure that unites previously separate devices: security alarms, smoke detectors, access control systems and other modules that can communicate without the server. In practice, this allows for a variety of different scenarios. For example, a network of several cameras will be able to accurately track a suspicious machine, passing the baton from one observation point to another and thus forming a complete picture of what is happening. Another scenario: if one of the cameras is disabled by the attacker, the neighbors automatically take control of the blind zone, and an alarm signal will be sent to the operator’s console. Interfacing with smoke detectors makes it possible to instantly assess the fire area, detect the fact of arson or false alarms, and integration with the access control system allows you to quickly record attempts at unauthorized entry to the protected object and establish the identity of the criminal.
Often, such systems even help to prevent an offense. For example, take a simple four-component scheme that operates without server participation. For the protection of the perimeter is used thermal imager. When an abnormal situation occurs (an intruder penetrates an object at night), the imager sends a signal to a PTZ camera, which turns in the specified direction. At the same time, the outside light is turned on and the loudspeaker is activated, broadcasting a pre-recorded message demanding to leave the protected area immediately. According to Alarm.org, this approach can prevent up to 74% of illegal invasions.
The appearance of additional functionality and improvement of technical characteristics led to a significant expansion of the scope of application of video surveillance systems. One of the most promising areas, however strange it may sound, was retail. Analysts at Transparency Market Research consider retailers to be one of the main drivers for the further development of the industry, because here video analytics systems allow solving various marketing tasks.
A typical scenario is the determination of the conversion rate based on information about the number of visitors and the number of customers who passed through the cash counters. If you add a face recognition system here, integrating it with the existing loyalty program, it becomes possible to study the behavior of customers with reference to socio-demographic factors for the subsequent formation of personalized offers (individual discounts, bundles at a reasonable price, etc.). Simpler use cases are no less effective, helping to save money on staff and improve customer experience. For example, the Chinese mega-corporation Baidu successfully refused not only to sell tickets to its amusement parks, but also from the ticket collectors themselves. Henceforth, facial recognition terminals are installed at the checkpoint, taking pictures of visitors and comparing them with the database of customers who have subscribed.
Baidu has replaced ticket collectors with pillars
The whole procedure takes no more than 0.6 seconds, and the recognition accuracy reaches 99.77%! Alipay users even had the opportunity to pay for online purchases using selfies, and all thanks to the integration of the payment service with the Face ++ visual identification platform created by Megvii. Currently, it is also successfully used by the Chinese division of UBER.
High technology opens up new horizons in the fight against petty theft in supermarkets. It is extremely difficult to resist this phenomenon, since professional thieves act as part of organized groups whose members play special roles (theft itself, the distraction of the guards, cover, etc.). The system of face recognition can automatically establish the fact that the same people regularly visit the store at the same time, even if they come separately, and transmit information about their presence to the guard post. If the suspicions are confirmed and it really turns out to be thieves, their faces will be entered into the retailer's database and an alarm will be triggered when intruders appear in any branch of the retail network.
On paper, everything looks good, but what about the implementation? If we talk about a facial recognition system like the one described above, its implementation is fraught with significant costs. It’s one thing to scan a face full face at a checkpoint, and quite another is in the stream, at different angles and in different lighting conditions. Here, only three-dimensional modeling of persons in real time using stereo cameras and algorithms of convolutional neural networks can demonstrate sufficient efficiency, which will inevitably require upgrading the existing infrastructure, as well as increasing data storage capacity. But many business representatives are not ready to go for additional expenses, even when it comes to classic security systems.
According to Denis Mikhailov, a senior expert at the video engineering department of the Expert Forensic Center of the Ministry of Internal Affairs of the Russian Federation, the vast majority of video recordings are not suitable for conducting investigative actions precisely because of the desire of businessmen to save on data storage. A typical situation is the creation of a two-component registration system: the main server stores records for the last week in high resolution at a frequency of 25 frames per second, and only a pinched image with a frame rate of 5 frames per second is archived. Unfortunately, such a video can not help to identify the attacker, and in some cases can not even recognize the license plates. Sometimes it’s ridiculous: the victims donate the assembled DVRs for examination, because instead of modern NVR or NAS, such outdated equipment is used that information can be read only if there is an original device.
Some entrepreneurs go for outright fraud. The contract indicates the installation of PTZ video cameras of high definition, and in fact, cheap devices are mounted that write an image in the 800x600 format and do not support remote control. With a favorable set of circumstances, such systems can work for years, and the owners will find out about their complete inconsistency only in the case of force majeure, when the outdated infrastructure cannot cope with the tasks set.
However, even if we ignore the conscious abuses and stinginess, related industries suffer from completely objective problems. One of the promising areas of video analytics development is the use of machine vision in industrial production.
The use of machine vision in industry
The technology can be used to detect product defects, measure cargo dimensional parameters, monitor employee admission and for other purposes. In the future, this would allow putting on stream the creation of fully automated production complexes capable of operating under the supervision of several operators. However, according to information from the same Transparency Market Research, the growth rate of this market segment turns out to be much more modest: the forecast for 2022 is $ 28.5 billion (compared to $ 71 billion in the security sector). If we recall that in 2014 this figure was 15.7 billion, the output is a gain of less than 2 times in eight years.
There are exactly three deterrent factors. The first is the lack of specialists with the necessary skill set to manage such systems. The second is high demands on the quality and security of equipment. The third is the lack of unified industry standards. “When it comes to security, versatility becomes one of the important advantages of video surveillance systems. Whatever the protected object (office, warehouse, supermarket, hotel), the pool of tasks to be solved is identical, that is, once developed hardware can be used anywhere, the additional functionality can be implemented at the software level. At the same time, each industrial enterprise is characterized by unique production processes, and automation tasks can no longer be solved by standard means - each of them requires a unique development, ”commented the current situation on Macroscop CEO Artem Razumov.
Indeed, even the most complex of the above scenarios for retail needs can be implemented on the basis of the same equipment used in security systems, because its capabilities depend primarily on the potential of video analytics algorithms and their proper use. However, the infrastructure created at the manufacturing site has a lot of additional requirements - ranging from the stability of its components to the effects of aggressive environmental factors (relevant for heavy industry) and the creation of unique quality control devices for microcircuits, welds or motor units. Although there is nothing impossible in the development of the latter, the client himself will finance it, since, due to the specificity of the periphery, it is unlikely to be used elsewhere. As a result, such projects turn out to be profitable only for large corporations, while all the rest still have to rely on cheap labor.
If the introduction of machine vision systems in an enterprise is a really difficult task, and the very possibility of developing such an infrastructure is seriously affected by macroeconomic factors, then when it comes to small and medium businesses, the problem of limited budget can be solved by optimizing and scaling existing ones. capacities. For example, increasing the density of data storage will help to save on the rental of space allocated for the data center, and the use of “cooler” disks will reduce the operating costs of electricity and air conditioning of the premises.
Considering all of the above, Western Digital has developed the concept of Core to Edge storage for Surveillance, offering customers a comprehensive set of advanced solutions for camera-to-server video recording systems. The combination of advanced technologies, reliability, capacity and performance allows you to build a harmonious ecosystem that can solve almost any task and optimize the cost of deployment and maintenance of infrastructure. The flagship line of our company is a family of specialized hard drives for WD Purple video surveillance systems, globally updated in 2017. Among the main changes, it is necessary to indicate the appearance of a new 10 GB hard drive, adding support for simultaneous work with 64 high-definition cameras, as well as the ability to work as part of servers and NAS designed for installing 16 disks.
Comparison of WD Purple of different generations
All of the above was made possible through the use of the following technological solutions:
HelioSeal. The older models of the WD Purple line with a capacity of 8 and 10 TB are made on the basis of the HelioSeal platform. The cases of these accumulators are absolutely hermetic, and the internal cavities, as the name implies, are filled with rarefied helium. Reducing the resistance of the gaseous medium and turbulence indicators made it possible to reduce the thickness of the magnetic plates, as well as to achieve a higher PMR recording density due to improved head positioning accuracy (using Advanced Format Technology). As a result, the transition to WD Purple provides an increase in data storage capacity up to 33% without the need to scale the infrastructure (in the same racks). In addition, helium storage devices are 26% more energy efficient than conventional HDDs by reducing the power consumption required for spinning and spindle rotation. Additional savings are in reducing the cost of air conditioning: with the same load, WD Purple is colder than its counterparts, on average, by 5 ° C.
AllFrame. The slightest interruption in recording can lead to the loss of critical video data, making it impossible to further analyze the information received. To prevent this, support for the optional section Streaming Feature Set of the ATA protocol was introduced into the firmware of the “purple” series disks. Among its capabilities, it is necessary to highlight the optimization of the use of cache depending on the number of video streams being processed and managing the priority of execution of read / write commands, thereby minimizing the probability of frame skipping and the appearance of image artifacts. In turn, the innovative set of algorithms AllFrame provides the ability to use hard drives in systems that process a significant number of isochronous streams: WD Purple drives support simultaneous work with 64 high-definition cameras.
Tler. The key tasks that have to be solved when designing video surveillance systems are to increase the uptime and reduce downtime due to malfunction. One of the most common problems when working with high-load servers is the spontaneous decomposition of a RAID array with redundancy caused by exceeding the allowed error correction time. The Time Limited Error Recovery option helps to avoid shutting down the HDD in case the timeout goes beyond 7 seconds: to prevent this, the drive will send a corresponding signal to the RAID controller, after which the correction procedure will be delayed until the system is idle.
RAFF. WD Purple with a capacity of 4, 6, 8 and 10 TB acquired the mechanism of adaptive compensation of vibration Rotary Acceleration Feed Forward (RAFF), which reduced the performance loss under the influence of rotational vibration to a record 4% (in conventional desktop hard drives it reaches 76%). This was made possible by calculating the compensatory trajectory of movement of the writing heads in real time based on the differential signals received from the built-in accelerometers, which record both linear and angular acceleration.
In addition, all models of the updated line have received a special coating that protects the printed circuit board from corrosion and the effects of aggressive chemical agents, which provides additional protection for both the devices themselves and the stored information.
After the acquisition of SanDisk, the Western Digital portfolio was supplemented with SanDisk Industrial memory cards with capacities ranging from 8 GB to 128 GB, aimed at use in car and personal video recorders, as well as in stationary cameras (as backup storage or buffer memory). Among their features, it is necessary to highlight the support of the error correction function ECC, which guarantees data integrity even during power outages, as well as the wear leveling algorithm, which helps to significantly increase the working life of the product. The ability to bind to the host and OTP-identifier exclude damage or theft of information during MITM attacks in IP video surveillance systems. Finally, a wide range of operating temperatures (from –40 to 85 ° C) allows the drives to be successfully used as part of surveillance cameras and even in industrial production conditions.
The world's first video surveillance system was developed by the Siemens concern in 1942 specifically for the Peenemünde test site, where Fau-2 missiles were tested. It was designed by the engineer Walter Bruch - the “father” of the PAL analog color television standard, adopted in 1966 as a state television in Germany, the United Kingdom and other Western European countries. With all the advantages, this solution had one major drawback: the operator could not leave the workplace for a second, as the signal was not recorded in any way. This continued until 1951, when video recorders (Video Tape Recorder, VTR) appeared, created by Ampex under the direction of electrical engineer Alexander Ponyatov, who emigrated to the United States in the late 1920s.
The first Ampex VR 1000B VCR
The importance of this invention for the industry is difficult to overestimate, because, according to IMS Research, even a physically healthy, well-rested person begins to miss up to 45% of critical events occurring on the monitor after 12 minutes of work, and after 10 minutes, this figure tends to a mark of 95%. Given this fact, the IP boom that occurred in 2014 does not seem so strange. Despite the emergence of AHD (Analog High Definition - High Definition Analog Video), network cameras turned out to be much more popular than analog solutions that were common at the time. And it's not just the picture quality. The main thing here is the vast opportunities that open up digital devices in front of their owners. After all, if before the analysis of the information received was carried out only after the fact, today security systems allow solving a wide range of tasks in real time. Modern IP cameras are strikingly different from their progenitor Axis Neteye 200, created back in 1996.
')
IP camera Axis Neteye 200
Many models of network cameras are already completely self-sufficient, and the IoT concept provides the ability to create a combined infrastructure that unites previously separate devices: security alarms, smoke detectors, access control systems and other modules that can communicate without the server. In practice, this allows for a variety of different scenarios. For example, a network of several cameras will be able to accurately track a suspicious machine, passing the baton from one observation point to another and thus forming a complete picture of what is happening. Another scenario: if one of the cameras is disabled by the attacker, the neighbors automatically take control of the blind zone, and an alarm signal will be sent to the operator’s console. Interfacing with smoke detectors makes it possible to instantly assess the fire area, detect the fact of arson or false alarms, and integration with the access control system allows you to quickly record attempts at unauthorized entry to the protected object and establish the identity of the criminal.
Often, such systems even help to prevent an offense. For example, take a simple four-component scheme that operates without server participation. For the protection of the perimeter is used thermal imager. When an abnormal situation occurs (an intruder penetrates an object at night), the imager sends a signal to a PTZ camera, which turns in the specified direction. At the same time, the outside light is turned on and the loudspeaker is activated, broadcasting a pre-recorded message demanding to leave the protected area immediately. According to Alarm.org, this approach can prevent up to 74% of illegal invasions.
Alternative scenarios for the use of modern video surveillance systems
The appearance of additional functionality and improvement of technical characteristics led to a significant expansion of the scope of application of video surveillance systems. One of the most promising areas, however strange it may sound, was retail. Analysts at Transparency Market Research consider retailers to be one of the main drivers for the further development of the industry, because here video analytics systems allow solving various marketing tasks.
A typical scenario is the determination of the conversion rate based on information about the number of visitors and the number of customers who passed through the cash counters. If you add a face recognition system here, integrating it with the existing loyalty program, it becomes possible to study the behavior of customers with reference to socio-demographic factors for the subsequent formation of personalized offers (individual discounts, bundles at a reasonable price, etc.). Simpler use cases are no less effective, helping to save money on staff and improve customer experience. For example, the Chinese mega-corporation Baidu successfully refused not only to sell tickets to its amusement parks, but also from the ticket collectors themselves. Henceforth, facial recognition terminals are installed at the checkpoint, taking pictures of visitors and comparing them with the database of customers who have subscribed.
Baidu has replaced ticket collectors with pillars
The whole procedure takes no more than 0.6 seconds, and the recognition accuracy reaches 99.77%! Alipay users even had the opportunity to pay for online purchases using selfies, and all thanks to the integration of the payment service with the Face ++ visual identification platform created by Megvii. Currently, it is also successfully used by the Chinese division of UBER.
High technology opens up new horizons in the fight against petty theft in supermarkets. It is extremely difficult to resist this phenomenon, since professional thieves act as part of organized groups whose members play special roles (theft itself, the distraction of the guards, cover, etc.). The system of face recognition can automatically establish the fact that the same people regularly visit the store at the same time, even if they come separately, and transmit information about their presence to the guard post. If the suspicions are confirmed and it really turns out to be thieves, their faces will be entered into the retailer's database and an alarm will be triggered when intruders appear in any branch of the retail network.
Deterrent to the development of video surveillance systems
On paper, everything looks good, but what about the implementation? If we talk about a facial recognition system like the one described above, its implementation is fraught with significant costs. It’s one thing to scan a face full face at a checkpoint, and quite another is in the stream, at different angles and in different lighting conditions. Here, only three-dimensional modeling of persons in real time using stereo cameras and algorithms of convolutional neural networks can demonstrate sufficient efficiency, which will inevitably require upgrading the existing infrastructure, as well as increasing data storage capacity. But many business representatives are not ready to go for additional expenses, even when it comes to classic security systems.
According to Denis Mikhailov, a senior expert at the video engineering department of the Expert Forensic Center of the Ministry of Internal Affairs of the Russian Federation, the vast majority of video recordings are not suitable for conducting investigative actions precisely because of the desire of businessmen to save on data storage. A typical situation is the creation of a two-component registration system: the main server stores records for the last week in high resolution at a frequency of 25 frames per second, and only a pinched image with a frame rate of 5 frames per second is archived. Unfortunately, such a video can not help to identify the attacker, and in some cases can not even recognize the license plates. Sometimes it’s ridiculous: the victims donate the assembled DVRs for examination, because instead of modern NVR or NAS, such outdated equipment is used that information can be read only if there is an original device.
Some entrepreneurs go for outright fraud. The contract indicates the installation of PTZ video cameras of high definition, and in fact, cheap devices are mounted that write an image in the 800x600 format and do not support remote control. With a favorable set of circumstances, such systems can work for years, and the owners will find out about their complete inconsistency only in the case of force majeure, when the outdated infrastructure cannot cope with the tasks set.
However, even if we ignore the conscious abuses and stinginess, related industries suffer from completely objective problems. One of the promising areas of video analytics development is the use of machine vision in industrial production.
The use of machine vision in industry
The technology can be used to detect product defects, measure cargo dimensional parameters, monitor employee admission and for other purposes. In the future, this would allow putting on stream the creation of fully automated production complexes capable of operating under the supervision of several operators. However, according to information from the same Transparency Market Research, the growth rate of this market segment turns out to be much more modest: the forecast for 2022 is $ 28.5 billion (compared to $ 71 billion in the security sector). If we recall that in 2014 this figure was 15.7 billion, the output is a gain of less than 2 times in eight years.
There are exactly three deterrent factors. The first is the lack of specialists with the necessary skill set to manage such systems. The second is high demands on the quality and security of equipment. The third is the lack of unified industry standards. “When it comes to security, versatility becomes one of the important advantages of video surveillance systems. Whatever the protected object (office, warehouse, supermarket, hotel), the pool of tasks to be solved is identical, that is, once developed hardware can be used anywhere, the additional functionality can be implemented at the software level. At the same time, each industrial enterprise is characterized by unique production processes, and automation tasks can no longer be solved by standard means - each of them requires a unique development, ”commented the current situation on Macroscop CEO Artem Razumov.
Indeed, even the most complex of the above scenarios for retail needs can be implemented on the basis of the same equipment used in security systems, because its capabilities depend primarily on the potential of video analytics algorithms and their proper use. However, the infrastructure created at the manufacturing site has a lot of additional requirements - ranging from the stability of its components to the effects of aggressive environmental factors (relevant for heavy industry) and the creation of unique quality control devices for microcircuits, welds or motor units. Although there is nothing impossible in the development of the latter, the client himself will finance it, since, due to the specificity of the periphery, it is unlikely to be used elsewhere. As a result, such projects turn out to be profitable only for large corporations, while all the rest still have to rely on cheap labor.
Affordable and productive solutions for small businesses
If the introduction of machine vision systems in an enterprise is a really difficult task, and the very possibility of developing such an infrastructure is seriously affected by macroeconomic factors, then when it comes to small and medium businesses, the problem of limited budget can be solved by optimizing and scaling existing ones. capacities. For example, increasing the density of data storage will help to save on the rental of space allocated for the data center, and the use of “cooler” disks will reduce the operating costs of electricity and air conditioning of the premises.
Considering all of the above, Western Digital has developed the concept of Core to Edge storage for Surveillance, offering customers a comprehensive set of advanced solutions for camera-to-server video recording systems. The combination of advanced technologies, reliability, capacity and performance allows you to build a harmonious ecosystem that can solve almost any task and optimize the cost of deployment and maintenance of infrastructure. The flagship line of our company is a family of specialized hard drives for WD Purple video surveillance systems, globally updated in 2017. Among the main changes, it is necessary to indicate the appearance of a new 10 GB hard drive, adding support for simultaneous work with 64 high-definition cameras, as well as the ability to work as part of servers and NAS designed for installing 16 disks.
Comparison of WD Purple of different generations
All of the above was made possible through the use of the following technological solutions:
HelioSeal. The older models of the WD Purple line with a capacity of 8 and 10 TB are made on the basis of the HelioSeal platform. The cases of these accumulators are absolutely hermetic, and the internal cavities, as the name implies, are filled with rarefied helium. Reducing the resistance of the gaseous medium and turbulence indicators made it possible to reduce the thickness of the magnetic plates, as well as to achieve a higher PMR recording density due to improved head positioning accuracy (using Advanced Format Technology). As a result, the transition to WD Purple provides an increase in data storage capacity up to 33% without the need to scale the infrastructure (in the same racks). In addition, helium storage devices are 26% more energy efficient than conventional HDDs by reducing the power consumption required for spinning and spindle rotation. Additional savings are in reducing the cost of air conditioning: with the same load, WD Purple is colder than its counterparts, on average, by 5 ° C.
AllFrame. The slightest interruption in recording can lead to the loss of critical video data, making it impossible to further analyze the information received. To prevent this, support for the optional section Streaming Feature Set of the ATA protocol was introduced into the firmware of the “purple” series disks. Among its capabilities, it is necessary to highlight the optimization of the use of cache depending on the number of video streams being processed and managing the priority of execution of read / write commands, thereby minimizing the probability of frame skipping and the appearance of image artifacts. In turn, the innovative set of algorithms AllFrame provides the ability to use hard drives in systems that process a significant number of isochronous streams: WD Purple drives support simultaneous work with 64 high-definition cameras.
Tler. The key tasks that have to be solved when designing video surveillance systems are to increase the uptime and reduce downtime due to malfunction. One of the most common problems when working with high-load servers is the spontaneous decomposition of a RAID array with redundancy caused by exceeding the allowed error correction time. The Time Limited Error Recovery option helps to avoid shutting down the HDD in case the timeout goes beyond 7 seconds: to prevent this, the drive will send a corresponding signal to the RAID controller, after which the correction procedure will be delayed until the system is idle.
RAFF. WD Purple with a capacity of 4, 6, 8 and 10 TB acquired the mechanism of adaptive compensation of vibration Rotary Acceleration Feed Forward (RAFF), which reduced the performance loss under the influence of rotational vibration to a record 4% (in conventional desktop hard drives it reaches 76%). This was made possible by calculating the compensatory trajectory of movement of the writing heads in real time based on the differential signals received from the built-in accelerometers, which record both linear and angular acceleration.
In addition, all models of the updated line have received a special coating that protects the printed circuit board from corrosion and the effects of aggressive chemical agents, which provides additional protection for both the devices themselves and the stored information.
After the acquisition of SanDisk, the Western Digital portfolio was supplemented with SanDisk Industrial memory cards with capacities ranging from 8 GB to 128 GB, aimed at use in car and personal video recorders, as well as in stationary cameras (as backup storage or buffer memory). Among their features, it is necessary to highlight the support of the error correction function ECC, which guarantees data integrity even during power outages, as well as the wear leveling algorithm, which helps to significantly increase the working life of the product. The ability to bind to the host and OTP-identifier exclude damage or theft of information during MITM attacks in IP video surveillance systems. Finally, a wide range of operating temperatures (from –40 to 85 ° C) allows the drives to be successfully used as part of surveillance cameras and even in industrial production conditions.
Source: https://habr.com/ru/post/350542/
All Articles