![<?echo $_SERVER['SERVER_NAME'];?>](/template/twentyseventeen/skin/images/header.jpg)
In the second decade of the 21st century, public security monitoring is undoubtedly a hot topic of digitization. Its application has also penetrated into all aspects of society and has begun to influence and change people's lifestyle. Security monitoring takes into consideration public safety, industry and commerce, transportation, and home and consumption. Applications for public safety have covered urban area monitoring, government homeland security, and public security departments that are closely related to ensuring social stability; applications for industry and commerce have been expanded from key facilities to conventional places, making people in banks, games, and retail Activities in frequent areas are guaranteed to be safe; applications for transportation are continually being developed and new security measures have been put in place not only at stations, terminals, airports, but also at roads, junctions and fairways.
These applications mentioned above are only aimed at public safety. Video surveillance for home and consumer groups will be more widely promoted and popularized for personal safety.
In the public security monitoring and control application, the most traditional one is the access control, and only afterwards there is a fire detection device. Video surveillance was first used with closed-circuit television systems and then developed into analog video systems, and scaled up by the popularity of analog cameras and analog video matrices. Digital video surveillance is accompanied by the development of digital video compression technology, and it stands out in the competition with analog systems and establishes the mainstream application direction. The combination of a computer and network technology platform has greatly promoted the development of digital video technology, and has exerted tremendous potential in a timely manner. Digital video surveillance in hard disk storage, remote transmission, flexible networking, video analysis, etc., its superiority has not been compared to analog systems, and greatly squeezed the existence of analog applications. However, in the last decade, digital video was mainly standard definition and did not completely surpass the image quality of analog systems. In the new era, what are the advantages of digital video surveillance?
The author believes that the new trend of digital video surveillance can be summarized in three aspects, that is, to achieve high-definition, enhance intelligence and play an optical element. Only high-definition video can finally eliminate analog video. Only intelligence can replace humans. Only optical elements can substantially improve video effects.
To achieve high-definition for the analog system, it is necessary to talk about the clarity of the TV, PAL TV nominal number of vertical scanning lines is 625 lines, in addition to the 50 line retracement, the actual effective vertical resolution of 575 lines. Of course, different analog video and video equipment will be different. The computer is imaged on a monitor, and the resolution expressed is the number of horizontal pixels and the number of vertical pixels. Such as VGA (640 × 480), SVGA (800 × 600), XGA (1024 × 768) and so on.
The initial digitized video is CIF resolution, 320x240, and the compressed format is MPEG1, and the image quality still stays at the VCD level. With the compression format going up to MPEG4 and then to H. 264, compression efficiency has been greatly improved, and the quality of the picture is also good, but the resolution is still not enough to be compared with the simulation, can be displayed on a TV or computer, but there is no real requirement for clarity.
The resolution of D1, ie 720x480, does give the digital compression effect a refreshing feeling. The compressed format starts with MPEG2, which is the level of DVD, but the bit rate is high, so the widespread adoption in surveillance also requires MPEG4 or H. . 264 compression format, used to call D1 SD. If the DVD is made using NTSC, the pixels are 720×480, PAL pixels are 720×576, and the horizontal resolution of both is 720×3/4=540. The former is higher than the vertical resolution (480 lines). Slightly lower than vertical resolution (576 lines).
The resolution to really open and simulate the gap between image quality is 720P, which is 1280x720, which can also be considered as HD. What is the real HD? This is the so-called 1080P, which is 1920x1080. Figure 2 shows the comparison of horizontal and vertical pixels from SD to HD.
The normal frame rate from SD to HD is 30 frames per second, so H. 2641080p@30fps means to take H. The 264 format compresses 30 images of 1920x1080 pixels per second, which is a level that must be achieved to achieve high definition. On this basis, support for multiple video compression formats is also very important, such as MPEG4, MJPEG, VC1, and so on. This performance is equivalent to twice the 720P, so it can improve video quality and add more detail. 720P is quasi-high definition, if increase the frame rate to 30 frames per second, that is H. 264 & MPEG-4 @ 720p60, the image's clarity will be effectively improved. Of course, the high-definition monitor that highlights the picture will also use MJPEG@5Mps, which can achieve an image of 5 million pixels per second.
Currently, there are many manufacturers in the industry that can provide high-definition video compression single-chip system (SOC). The Texas Instruments (TI) provides a platform that can support various high-definition needs and supports multiple high-definition video compression formats. Higher system flexibility. The DM6467, which has been mass-produced, is a monolithically realizable H. 264 and other video compression formats High-definition or higher SOC, which is supported by the 600MHz clock version. The performance supported by the 264@1080P30, 1GHz clock version will nearly double. The platform integrates ARM9 processor and video coprocessor and high-speed C64+x DSP core, as well as BT1120 high-definition video input and output ports, in addition to a wealth of peripheral interfaces. In addition to HD video encoding and decoding, its DSP unit can take on intelligent video processing functions. High-definition video transcoding is also an important application direction of DM6467.
The same production capacity of DM365 supports H. 264@720P30/MPEG-4@1080i30 or, MPEG-4@1080P30/H. 264@1080P12 has already been used as a high-definition entry product in surveillance IP cameras.
The DM36x platform to be newly introduced is H. 264&MPEG-4@1080P30, and compatible with DM365 architecture, software and pins. This will be a truly full support for high-definition SOCs, with an integrated 400MHz ARM9 processor, video coprocessor, video processing subsystem, and rich peripheral interfaces as shown in Figure 4. The optical elements involved in the video processing subsystem are left to the following specific analysis.
The market is expecting a higher performance HD platform, H. The 264@1080P60 specification is not out of reach, and even one chip can support two channels. TI hopes to bring greater surprises to the digital video surveillance industry in the first year of the new era.
Enhancing intelligent intelligent video analysis should be eternal topic of digital video surveillance. The analog system can only handle the anomaly by artificially monitoring or repeatedly watching the videotape. Even with digital video recording, it is still difficult to get rid of human participation observation and recognition until video content analysis algorithms emerge. Early video content analysis based on the back-end processing of PCs can help people reduce the burden of heavy viewing, but it is not possible to discover and interfere with emergencies on the monitoring site in time. The introduction of DSP can make embedded video content analysis possible, so as shown in Figure 3, smart video can be fully applied in the security market to refine the product.
Video Surveillance Conventional Products DVRs, DVSs, and IP cameras can all add risk management and intelligent services by adding front-end intelligent video processing. In camera-related applications, intelligence can meet motion detection, centralized detection, and automatic tracking. demand.
In traditional access control systems, biometric recognition is not just a fingerprint, but it can also be face monitoring and recognition, which can automatically determine the identity and eliminate human disadvantages. For intrusion detection and alerting, the use of intelligent video means will be more flexible and user-friendly, so the video doorbell products will have the opportunity to give more features. In the application of fire prevention and life safety, the combination of smart video and traditional sensing methods will be more effective.
Its essence can make the application scope of video still have a lot of room for expansion. In combination with the trend of high-definition video surveillance, in the face of a large amount of video data storage or transmission, it will definitely take up a lot of system and equipment resources if it can be used. The means of intelligent video analysis will be able to effectively filter out the impact of the original video, retain the most valuable content and store or forward it in high definition.
In the global environment of stressing the "low-carbon" green, smart video management of air-conditioning can also be regarded as an innovative application. On the one hand, the intelligence of video surveillance is to ensure the normal flow of personnel in the workplace. On the other hand, it is also possible to detect changes in the number of people in various regions. It is not difficult to determine how the appropriate temperature in the room will be controlled. Or cooling.
Based on this way of thinking, more intelligent video applications can be imagined. Therefore, enhanced intelligence will be the only way for digital video surveillance to enhance the value space in the current severe market competition.
The most solid processing platform for digital video is DSP. TI's star platform, the DM642, plays an important role in smart video embedding applications. The subsequent Davinci series platforms can meet the needs of digital video in different areas and continue to gain market share. success.
In order for intelligent video applications to continue to grow strongly, roadmap planning and development ecosystem (ECOSystem) for DSP platforms are critical. TI's successful introduction from C64+ to the widespread adoption of C64+x will also integrate the C674x core into a new generation of high-definition SOCs. This is a floating-point and fixed-point compatible DSP core that will be more beneficial for more intelligent video implementations.
To promote the development of intelligent video, TI developed and assembled more than fifty kinds of software kernels to form an intelligent video elementary library VLIB based on the C64+x platform, including background modeling and background extraction algorithms, target feature extraction, tracking and identification, and low-level Pixel processing and so on. VLIB is a highly scalable software foundation platform for applications such as video analytics, computer and automotive vision, embedded vision systems, and consumer electronics vision. VLIB can be registered and downloaded from TI's public website. There are hundreds of customers now using TI's DSP for intelligent video software development, which will also become an inexhaustible source of innovation for digital video surveillance.
In the process of replacing the analog system with an optical element digital video system, there is almost the same analog device that has been used for a long time. This is the analog camera (CCTV Camera). The reason is that the camera seems to be an integral part of the optoelectronics, and the optical part thereof Have been fixed. In fact, the integrated camera consists of two parts, the front end is a CCD or CMOS video sensor, the back end is a video signal processor (ISP) for processing the original video signal, and then output as a BT656 video signal. In the past, ISPs were mastered by sensor manufacturers, and it was difficult for customers to develop. So is this situation likely to change? This relies on the "optical elements" integrated in the SOC processor.
In order to make the digital video system more distinctive, especially to meet the needs of different scenes, the understanding and mastery of the optical elements becomes more and more important, especially the integrated IP camera. TI's DM36x was born out of a digital camera platform, and the performance of the video processing subsystem (VPSS) is particularly critical. Figure 4 shows the functional block diagram of the DM36x video processing subsystem. Included in the VPSS is a CCD or CMOS controller that directly receives the raw video signal, which is then pre-processed by the integrated ISP. Among these are the so-called H3A processing, which is Auto Exposure (AE), Auto White Balance (AWB), Auto Focus (AF). ). Of course, achieving such a process requires a rich professional optical experience to be able to obtain the ideal image effect.
The so-called optical element also has lens aberration correction mode (LDC) processing. VPSS also includes hardware face detection engine, image coprocessor (IMCOP), image processing accelerator (IPIPE), and enhanced screen overlay (OSD). VPSS also has preview and zoom capabilities that support high-definition component output. TI will adopt a new generation of ISP in the new SOC and add more abundant optical elements.
The optical element will be the favorable guarantee for future comprehensive digital high-definition and intelligentization of digital video. The reason for this lies in the fact that a large number of high-definition video sensors are of CMOS type. High-definition integrated video products will be the direction of video surveillance vendors' R&D and production. In addition, the application of intelligent monitoring products is closely coupled with the optical factors of the environment. Only by truly mastering the adaptive application of optics can we expand various intelligent video applications more effectively.
In short, the new era will surely further advance the field of intelligent high-definition video surveillance. China has become the fastest growing market for video surveillance in the world, and its innovative momentum and potential are attracting attention. By seizing the opportunity of high-definition and intelligence, combined with the profound optical skills China has accumulated, Chinese security companies will have the ability to enhance their advantages and benefit the world.
These applications mentioned above are only aimed at public safety. Video surveillance for home and consumer groups will be more widely promoted and popularized for personal safety.
In the public security monitoring and control application, the most traditional one is the access control, and only afterwards there is a fire detection device. Video surveillance was first used with closed-circuit television systems and then developed into analog video systems, and scaled up by the popularity of analog cameras and analog video matrices. Digital video surveillance is accompanied by the development of digital video compression technology, and it stands out in the competition with analog systems and establishes the mainstream application direction. The combination of a computer and network technology platform has greatly promoted the development of digital video technology, and has exerted tremendous potential in a timely manner. Digital video surveillance in hard disk storage, remote transmission, flexible networking, video analysis, etc., its superiority has not been compared to analog systems, and greatly squeezed the existence of analog applications. However, in the last decade, digital video was mainly standard definition and did not completely surpass the image quality of analog systems. In the new era, what are the advantages of digital video surveillance?
The author believes that the new trend of digital video surveillance can be summarized in three aspects, that is, to achieve high-definition, enhance intelligence and play an optical element. Only high-definition video can finally eliminate analog video. Only intelligence can replace humans. Only optical elements can substantially improve video effects.
To achieve high-definition for the analog system, it is necessary to talk about the clarity of the TV, PAL TV nominal number of vertical scanning lines is 625 lines, in addition to the 50 line retracement, the actual effective vertical resolution of 575 lines. Of course, different analog video and video equipment will be different. The computer is imaged on a monitor, and the resolution expressed is the number of horizontal pixels and the number of vertical pixels. Such as VGA (640 × 480), SVGA (800 × 600), XGA (1024 × 768) and so on.
The initial digitized video is CIF resolution, 320x240, and the compressed format is MPEG1, and the image quality still stays at the VCD level. With the compression format going up to MPEG4 and then to H. 264, compression efficiency has been greatly improved, and the quality of the picture is also good, but the resolution is still not enough to be compared with the simulation, can be displayed on a TV or computer, but there is no real requirement for clarity.
The resolution of D1, ie 720x480, does give the digital compression effect a refreshing feeling. The compressed format starts with MPEG2, which is the level of DVD, but the bit rate is high, so the widespread adoption in surveillance also requires MPEG4 or H. . 264 compression format, used to call D1 SD. If the DVD is made using NTSC, the pixels are 720×480, PAL pixels are 720×576, and the horizontal resolution of both is 720×3/4=540. The former is higher than the vertical resolution (480 lines). Slightly lower than vertical resolution (576 lines).
The resolution to really open and simulate the gap between image quality is 720P, which is 1280x720, which can also be considered as HD. What is the real HD? This is the so-called 1080P, which is 1920x1080. Figure 2 shows the comparison of horizontal and vertical pixels from SD to HD.
The normal frame rate from SD to HD is 30 frames per second, so H. 2641080p@30fps means to take H. The 264 format compresses 30 images of 1920x1080 pixels per second, which is a level that must be achieved to achieve high definition. On this basis, support for multiple video compression formats is also very important, such as MPEG4, MJPEG, VC1, and so on. This performance is equivalent to twice the 720P, so it can improve video quality and add more detail. 720P is quasi-high definition, if increase the frame rate to 30 frames per second, that is H. 264 & MPEG-4 @ 720p60, the image's clarity will be effectively improved. Of course, the high-definition monitor that highlights the picture will also use MJPEG@5Mps, which can achieve an image of 5 million pixels per second.
Currently, there are many manufacturers in the industry that can provide high-definition video compression single-chip system (SOC). The Texas Instruments (TI) provides a platform that can support various high-definition needs and supports multiple high-definition video compression formats. Higher system flexibility. The DM6467, which has been mass-produced, is a monolithically realizable H. 264 and other video compression formats High-definition or higher SOC, which is supported by the 600MHz clock version. The performance supported by the 264@1080P30, 1GHz clock version will nearly double. The platform integrates ARM9 processor and video coprocessor and high-speed C64+x DSP core, as well as BT1120 high-definition video input and output ports, in addition to a wealth of peripheral interfaces. In addition to HD video encoding and decoding, its DSP unit can take on intelligent video processing functions. High-definition video transcoding is also an important application direction of DM6467.
The same production capacity of DM365 supports H. 264@720P30/MPEG-4@1080i30 or, MPEG-4@1080P30/H. 264@1080P12 has already been used as a high-definition entry product in surveillance IP cameras.
The DM36x platform to be newly introduced is H. 264&MPEG-4@1080P30, and compatible with DM365 architecture, software and pins. This will be a truly full support for high-definition SOCs, with an integrated 400MHz ARM9 processor, video coprocessor, video processing subsystem, and rich peripheral interfaces as shown in Figure 4. The optical elements involved in the video processing subsystem are left to the following specific analysis.
The market is expecting a higher performance HD platform, H. The 264@1080P60 specification is not out of reach, and even one chip can support two channels. TI hopes to bring greater surprises to the digital video surveillance industry in the first year of the new era.
Enhancing intelligent intelligent video analysis should be eternal topic of digital video surveillance. The analog system can only handle the anomaly by artificially monitoring or repeatedly watching the videotape. Even with digital video recording, it is still difficult to get rid of human participation observation and recognition until video content analysis algorithms emerge. Early video content analysis based on the back-end processing of PCs can help people reduce the burden of heavy viewing, but it is not possible to discover and interfere with emergencies on the monitoring site in time. The introduction of DSP can make embedded video content analysis possible, so as shown in Figure 3, smart video can be fully applied in the security market to refine the product.
Video Surveillance Conventional Products DVRs, DVSs, and IP cameras can all add risk management and intelligent services by adding front-end intelligent video processing. In camera-related applications, intelligence can meet motion detection, centralized detection, and automatic tracking. demand.
In traditional access control systems, biometric recognition is not just a fingerprint, but it can also be face monitoring and recognition, which can automatically determine the identity and eliminate human disadvantages. For intrusion detection and alerting, the use of intelligent video means will be more flexible and user-friendly, so the video doorbell products will have the opportunity to give more features. In the application of fire prevention and life safety, the combination of smart video and traditional sensing methods will be more effective.
Its essence can make the application scope of video still have a lot of room for expansion. In combination with the trend of high-definition video surveillance, in the face of a large amount of video data storage or transmission, it will definitely take up a lot of system and equipment resources if it can be used. The means of intelligent video analysis will be able to effectively filter out the impact of the original video, retain the most valuable content and store or forward it in high definition.
In the global environment of stressing the "low-carbon" green, smart video management of air-conditioning can also be regarded as an innovative application. On the one hand, the intelligence of video surveillance is to ensure the normal flow of personnel in the workplace. On the other hand, it is also possible to detect changes in the number of people in various regions. It is not difficult to determine how the appropriate temperature in the room will be controlled. Or cooling.
Based on this way of thinking, more intelligent video applications can be imagined. Therefore, enhanced intelligence will be the only way for digital video surveillance to enhance the value space in the current severe market competition.
The most solid processing platform for digital video is DSP. TI's star platform, the DM642, plays an important role in smart video embedding applications. The subsequent Davinci series platforms can meet the needs of digital video in different areas and continue to gain market share. success.
In order for intelligent video applications to continue to grow strongly, roadmap planning and development ecosystem (ECOSystem) for DSP platforms are critical. TI's successful introduction from C64+ to the widespread adoption of C64+x will also integrate the C674x core into a new generation of high-definition SOCs. This is a floating-point and fixed-point compatible DSP core that will be more beneficial for more intelligent video implementations.
To promote the development of intelligent video, TI developed and assembled more than fifty kinds of software kernels to form an intelligent video elementary library VLIB based on the C64+x platform, including background modeling and background extraction algorithms, target feature extraction, tracking and identification, and low-level Pixel processing and so on. VLIB is a highly scalable software foundation platform for applications such as video analytics, computer and automotive vision, embedded vision systems, and consumer electronics vision. VLIB can be registered and downloaded from TI's public website. There are hundreds of customers now using TI's DSP for intelligent video software development, which will also become an inexhaustible source of innovation for digital video surveillance.
In the process of replacing the analog system with an optical element digital video system, there is almost the same analog device that has been used for a long time. This is the analog camera (CCTV Camera). The reason is that the camera seems to be an integral part of the optoelectronics, and the optical part thereof Have been fixed. In fact, the integrated camera consists of two parts, the front end is a CCD or CMOS video sensor, the back end is a video signal processor (ISP) for processing the original video signal, and then output as a BT656 video signal. In the past, ISPs were mastered by sensor manufacturers, and it was difficult for customers to develop. So is this situation likely to change? This relies on the "optical elements" integrated in the SOC processor.
In order to make the digital video system more distinctive, especially to meet the needs of different scenes, the understanding and mastery of the optical elements becomes more and more important, especially the integrated IP camera. TI's DM36x was born out of a digital camera platform, and the performance of the video processing subsystem (VPSS) is particularly critical. Figure 4 shows the functional block diagram of the DM36x video processing subsystem. Included in the VPSS is a CCD or CMOS controller that directly receives the raw video signal, which is then pre-processed by the integrated ISP. Among these are the so-called H3A processing, which is Auto Exposure (AE), Auto White Balance (AWB), Auto Focus (AF). ). Of course, achieving such a process requires a rich professional optical experience to be able to obtain the ideal image effect.
The so-called optical element also has lens aberration correction mode (LDC) processing. VPSS also includes hardware face detection engine, image coprocessor (IMCOP), image processing accelerator (IPIPE), and enhanced screen overlay (OSD). VPSS also has preview and zoom capabilities that support high-definition component output. TI will adopt a new generation of ISP in the new SOC and add more abundant optical elements.
The optical element will be the favorable guarantee for future comprehensive digital high-definition and intelligentization of digital video. The reason for this lies in the fact that a large number of high-definition video sensors are of CMOS type. High-definition integrated video products will be the direction of video surveillance vendors' R&D and production. In addition, the application of intelligent monitoring products is closely coupled with the optical factors of the environment. Only by truly mastering the adaptive application of optics can we expand various intelligent video applications more effectively.
In short, the new era will surely further advance the field of intelligent high-definition video surveillance. China has become the fastest growing market for video surveillance in the world, and its innovative momentum and potential are attracting attention. By seizing the opportunity of high-definition and intelligence, combined with the profound optical skills China has accumulated, Chinese security companies will have the ability to enhance their advantages and benefit the world.