Video ABC / Abbreviations
AHD - Analog High Definition is both a general term and the name of a specific variant/ implementation of megapixel analog, in contrast to the more widely used megapixel IP.
AOV - Angle of View describes the camera's coverage area in degrees. It is often used interchangeably with field of view (FOV), though FOV typically refers to the scene's width, while AOV refers to angle.
Aspect Ratio - The ratio of image height to width. In surveillance, 16:9 is the most common aspect ratio as it is used by HD 720p and 1080p sensors, as well as most monitors in use today. 4:3 and 3:2 ratios are also in use in some cameras, though less commonly than previous.
AWG - American Wire Gauge is a standard scale of wire gauge used predominantly in North America, though in some other parts of the world. It is most commonly referred to when specifying low voltage power, e.g., "two-conductor 16 AWG" or in Category cable (such as Cat 5e or 6), which typically use 24 AWG conductors.
BLC - Back Light Compensation is a digital image adjustment which adjusts exposure to attempt to approve details on subjects in front of strong backlight. However, it uses a single exposure for the entire scene, so brighter areas may become washed out. It is not a substitute for WDR.
CBR - Constant Bit Rate is a camera streaming mode that aims for a constant bandwidth level while video quality is allowed to vary, which may cause image degradation in busy scenes if CBR bitrate is set too low. Contrast to VBR/MBR.
CCD - Charge Coupled Device, a type of image sensor which reads pixels of an image one by one. CCDs were previously the most popular sensor type used in surveillance but have been replaced by CMOS in the vast majority of cameras (see CMOS below).
CMOS - Complementary Metal Oxide Semiconductor. In surveillance, a CMOS sensor reads pixels of an image all at once. CMOS sensors are used in most cameras today, due to lower production cost and advances in low light, WDR and high-resolution performance.
CNN - Convolutional Neural Network. CNNs are used in deep learning to classify images, perform object recognition, and detect image similarity.
dB - Decibel. Decibels are a measurement of the difference in two power levels, most commonly used to specify WDR in surveillance (e.g., 58dB, 130dB, etc.), though with questionable accuracy and no independent validation. They are also used to measure audio levels, such as in public address or life safety applications.
Deep Learning - An algorithm which goes through a number of hierarchical classification stages, effectively filters, to reach decisions, most commonly human, vehicle, animal, etc., classifications in surveillance. Deep learning analytics are becoming increasingly common in 2020 and beyond.
DHCP - Dynamic Host Configuration Protocol assigns an IP address to devices on a network within a specified range (called a "scope"). Most IP cameras default to using DHCP to speed initial setup of devices.
DNR - Digital Noise Reduction uses software (typically on board cameras) to attempt to reduce image noise present in low light surveillance images. There are two key types of DNR, spatial (within a single frame) and temporal (comparing two or more frames) which are typically combined in what is known as "3D DNR." Applying excessive DNR may introduce motion blur, a problem on some modern models.
DVR - A Digital Video Recorder receives video from connected analog sources (typically SD or HD analog cameras), encodes it, and stores it to disk. Contrast this with an NVR (network video recorder) which captures network streams from IP cameras. Note that some DVRs are referred to as "hybrid" and may record both analog and IP cameras.
DWDR - "Digital" Wide Dynamic Range is essentially a contrast adjustment on multiple regions of the scene, instead of the overall image. Can sometimes be referred to as "Electronic" WDR, and is better in some scenes than doing nothing, but does not compare to "true" WDR performance.
EIS - Electronic Image Stabilization is a camera feature which attempts to minimize blurring and compensate for device vibration, such as cameras mounted on tall poles. This is a common feature in high end cameras today. Some third parties offer software or hardware based EIS add-ons, as well, though these are very rarely used.
EOL - End-Of-Life describes products which have been discontinued by the manufacturer. This means the product is no longer sold and support may be limited, either immediately or in some period of time.
F/F-Stop - F-stop measures the relative amount of light that a lens passes, expressed on a logarithmic scale (e.g. F1.4, F2.8, etc.). The lower the F number, the more light the lens passes. When lenses / cameras specify F numbers, they always provide the lowest F number that the lens supports when the iris is all the way open.
FPS - Frames Per Second. The number of images a camera captures and/or transmits in one second. 30 FPS is typically referred to as "full frame rate", while average is between 10-15 in surveillance.
FOV - Field of View. In simple terms, refers to what a camera can "see." However, there are many factors into how a FOV is calculated, and the impacts of lenses and sensor types further complicate how a FOV is measured and compared. Note that FOV is often used interchangeably used with Angle of View (AOV), but AOV is the angle, while FOV is the width.
Frame Types (I-Frame, P-Frame, B-frame)
I-frames, P-frames, and B-frames are different types of frames used in encoding surveillance (and other) video in the H.264, H.265, and MPEG-4 standards.
- I-Frame - An I-frame (or inter-coded picture) contains the entire image. In surveillance, this means that a camera will encode the entire image as is, and not the predicted changes as seen with P or B frames.
- P-Frame - A P-frame can be thought of as "predicted picture", which holds only the changes in the image from the previous frame. Take someone walking through a static scene for instance. Only the person's movement needs to be encoded, not the background.
- B-Frame - A "Bidirectional predicted picture" is similar to a P-frame. However, B-frames can use one, two, or more than two previously decoded pictures as references during decoding. This typically means fewer bits for encoding than either I or P-frames.
GOP / GOV - Group Of Pictures / Group Of Video. Refers to a collection of frames, beginning with an I-frame, including all subsequent P and B frames, until the next I-frame. For example, a GOP of 15 would include one I-frame, followed by 14 P and/or B frames, depending on structure.
HD-CVI - High Definition Composite Video Interface is an HD analog standard developed by Dahua for transmission of HD video over coax. Originally a proprietary technology, multiple manufacturers now offer "universal" DVRs or encoders which support AHD, CVI, and TVI.
HD-TVI - High Definition Transport Video Interface is an HD analog standard developed by chip manufacturer TechPoint and adopted by Hikvision, who remains the largest manufacturer of TVI product.
HDD - Hard Disk Drive. The most commonly used storage medium in surveillance, used in DVRs and NVRs.
HLC - High Light Compensation. A feature included in some IP cameras which attempts to reduce the effects of strong light sources, such as headlights, which would otherwise wash out the image.
IK - IK ratings are standard measurements of impact resistance defined in EN 62262, ranging from IK00 to IK10, though manufacturers have tested beyond IK10, as well, calling it IK10++. In surveillance, these ratings describe vandal resistance of cameras, with IK10 generally referred to as "vandal resistant", though there is no standard which requires this.
IP (Ingress Protection) - Ingress Protection (IP) ratings dictate how hardened a device (e.g., cameras is against dust, water and other debris or pressure. IP66 rating or higher is typically referred to as "weatherproof", though there is no standard for this terminology. IP67 and IP6K9K are ratings beyond IP66.
IPC - An Internet Protocol Camera, or IP Camera, is a digital surveillance camera that is able to send and receive information over a network. The term 'IPC' is used in some regions of the world as an acronym and often used in many manufacturers' part numbers.
IR - Infrared, light above the visible spectrum of humans, ranging from 700nm to 1000nm. IR is used in integrated and external illuminators, typically in the 850nm or 940nm wavelengths, to provide light which cameras can "see" but humans cannot.
LAN - Local Area Network, a collection of devices connected together in one physically location, e.g., an office or school. LAN is used to refer to one facility's network, in contrast with WAN (wide area network), which refers to a collection of LANs connected across some distance.
MAC Address - A Media Access Control address is a specific identifier assigned to a Network Interface Card, or NIC. It is a hardware identification number that uniquely identifies each device on a network.
Machine Learning - Uses pre-programmed instructions to allow a computer to recognize an image of an object. When a programmed algorithm is then fed video, it will look for these specified parameters. The difference with deep learning is the algorithm should come up with criteria on its own that is very similar to what the machine learning algorithm was manually programmed with.
MBR - Maximum Bit Rate, which is a streaming mode which allows the camera's bit rate to vary up to a specified maximum, often referred to as a cap, while maintaining a constant video quality level.
MP - Megapixel directly means "a million pixels". In surveillance, multi-megapixel cameras such as 1080p/2MP, 4MP, 5MP, 4K/8MP, etc., are common, and the "MP" quantifier details the resolution of a camera.
NAS - Network Attached Storage. In surveillance, NAS refers to a small device used to store video on the network. Multiple recorders (VMS/NVR) may record to a single NAS.
NDAA - The US National Defense Authorization Act of 2018 banned the use of major surveillance manufacturers (including Dahua, Hikvision and Huawei) in various applications. In August 2019, this started with the US federal government ban of purchasing such equipment. In August 2020, as currently proposed it is scheduled to ban the use of federal funding for such products as well as those using those products from selling to the US government.
NEMA - National Electrical Manufacturer Association. In surveillance, the term NEMA typically is used to refer to specific weather ratings of enclosures, such as a NEMA 4X enclosure, which may be used outdoors or in wet locations. It is related to, but not the same as, IP ratings (IP66, IP67, etc.).
NIC - Network Interface Card. A component which allows a device to connect to a network, typically Ethernet. This is typically integrated in modern PCs/servers and found in IP cameras.
NTP - Network Time Protocol. Used to synchronize the time of devices to a reference time source or server. NTP is important in surveillance as improper time stamps may impact whether surveillance video is admissible as evidence in a court of law.
NVR - A Network Video Recorder encodes and stores video from cameras on the network, in contrast to DVRs which record analog video from locally connected cameras. NVR is sometimes used to describe a VMS server, as well, though VMS’s typically use a standard operating system and third-party hardware while NVRs are typically proprietary hardware and software.
MBR - Maximum Bit Rate is a streaming mode that allows the camera's bit rate to vary but maintains a constant video quality level up to a certain maximum bandwidth level. This is the most commonly used bit rate control that benefits from allowing adjustment but limiting wasteful spikes.
ONVIF - Stands for Open Network Video Interface Forum. ONVIF is a trade organization founded by Axis, Bosch, and Sony in 2008 which has developed API specifications for integrating security products, and collected into a number of "Profiles" which contain specific sets of functionalities:
- Profile S: S is the oldest and most broadly supported. Covers video streaming and the basics of sending video from a camera to a VMS/recorder.
- Profile G: G supports video storage. This profile can support retrieving and sending stored video from and IP camera with on-board video to a VMS/recorder.
- Profile Q: Q was created to simply discovering cameras and improve security by eliminating default passwords. While officially adopted, has little manufacturer support.
- Profile T: T improves integration of H.265 codec, motion detection, analytics, and other events, as well as camera settings such as exposure, focus, contrast, etc.
PoE - Power Over Ethernet (PoE) is a standard for supplying low voltage power over the same cable used to transmit Ethernet data. PoE is supported and used, in practice, in almost all professional IP cameras and installations.
PPF (PPM) - Pixel Per Foot (Pixel Per Meter) is a metric of image quality, obtained by dividing horizontal pixel resolution (e.g., 1920 in a 1080p camera) by field of view width (e.g., 32'). PPF acts as a baseline to get a sense of potential performance but does not account for outside influences such as lenses, compression, scene brightness, low light performance, and others.
PTZ - Pan/Tilt/Zoom. PTZ cameras contain a motorized mechanism which allows it to move in multiple directions and change its angle of view. As the name implies, the cameras can move horizontally left and right, vertically up or down, and zoom the lens in or out. Overall usage has declined significantly in the past decade as multi-imagers and multi-megapixel cameras have expanded, but PTZs are still popular in some applications.
PTRZ - Pan/tilt/roll/zoom. PTRZ cameras are similar to PTZ cameras but include the ability to rotate the camera's lens clockwise and counterclockwise. Unlike PTZ models, however, this capability is not intended for tracking of subjects or tours, but instead as an installation feature. PTRZ cameras are slow relative to true PTZs and are not warrantied for full time PTZ use.
Q - Stands for quantization, a measurement of compression on a standard logarithmic scale from 0 to 51, with lower numbers meaning less compression, and thus higher image quality.
RAID - Redundant Array of Inexpensive Disks. It is used to connect multiple secondary storage devices and use them as a single storage space or media. RAID is broken into several levels with the goal of distributing data across several disks, and above RAID 0, can provide redundancy should a disk fail.
RMA - Return Merchandise Authorization is the process of returning a product to receive a refund, replacement or repair. Returning products normally require the user to speak with technical support and go through whatever troubleshooting steps are necessary before an RMA is issued.
RTSP - Real Time Streaming Protocol. Surveillance cameras utilize RTSP streams as a link between a camera and RTSP viewers such as VLC or VMS client software.
SAN - Storage Area Network. A SAN technically describes an entire network but is typically used to describe a network storage server, similar to simpler NAS (network attached storage), but generally larger and using different protocols for increased performance.
Smart Codec - Smart codecs dynamically adjust compression, I-frame interval (GOP), frame rate, and other streaming factors based on scene complexity in order to reduce required bandwidth. Smart codecs have had the greatest impact on network bandwidth of any technology in surveillance, with savings of 95% possible compared to standard H.264/H.265.
Smart IR - A term used to describe features which compensate for moving objects or scene changes in integrated IR cameras. Smart IR may mean IR power is automatically adjusted, dimming to prevent overexposure of near objects and returning to full power for those further away. However, it may simply refer to a faster exposure adjustment used to compensate in integrated IR models, or in some cases may be used to describe cameras which adjust their illumination angle based on lens zoom. There is no standard for this terminology.
SNMP - Simple Network Management Protocol is used for basic monitoring of devices over the network. It may be used to monitor bandwidth, OS service status, uptime, and more, and provide alerts to users of device downtime when using a network monitoring system.
Static (IP Address) - Static IP addresses on network devices mean a unique address is manually assigned to each camera. This results in IP devices retaining the same IP address should it become disconnected on a network. This is the recommended method of assigning IP addresses in most networks.
Thermal - Thermal surveillance cameras use infrared radiation to produce an image, meaning no light is required to produce an image. Additionally, they are able to produce images in adverse conditions such as smoke and fog. However, since they do not see visible light, thermal cameras do not reproduce color, and fine details between areas or objects of similar temperature are lost. Some thermal cameras include visible light imaging, and some uses of thermal are expanding to measuring human body temperature.
UPS - Uninterruptible Power Supply, a backup power system consisting of batteries and an inverter used to maintain power during outages. Battery run time may last for a few minutes up to a few hours, depending on size.
VCA - Video Content Analysis (or sometimes referred to as Video Content Analytics), most often simply referred to as "analytics." Analytics may refer to any of a number of rules, such as intrusion, people detection, LPR, etc. Analytics in new models are increasingly based on Deep Learning (see above) for increased accuracy.
VBR - Variable Bit Rate, a streaming mode that allows the camera's bit rate to vary but maintains a constant video quality level. Contrast to CBR and MBR.
VMD - Video Motion Detection. A term to describe generally less sophisticated, less accurate attempts to detect if motion is occurring, most often based simply on the intensity and number of pixels changing. It can be better than recording continuously but it generally detects lots of noise and other irrelevant activity. There are many attempts to improve VMD, most notably video analytics / VCA.
VMS - Video Management Software is installed on COTS (Commercial Off-The-Shelf) hardware in contrast to NVRs / DVRs where the video management software is bundled / sold with the hardware.
VOIP - Voice over Internet Protocol (IP) describes any of a number of methods which transmit voice via a network/the internet. In surveillance this is most commonly used regarding intercoms/door phones, a growing segment of specialty IP cameras.
VSaaS - Video Surveillance as a Service is a general term for cloud-based video surveillance, including both hosted (video stored in the cloud), managed video (video stored locally) and hybrid approaches. There is no standard for the term, but it is widely used to express the general approach.
- Hosted Video - A type of VSaaS where video is both stored ('hosted') and managed in the cloud. This eliminates having to keep video and recorders on-site but requires enough bandwidth to move video to the cloud. In consumer applications, this is increasingly common but in commercial, with large camera counts, this is less common.
- Managed Video - A type of VSaaS where video is 'managed' in the cloud but stored locally at the customer site. This is the most common type of VSaaS today as it allows for benefits of cloud management (easier remote access, firmware updates, etc.) without having to transfer all video to the cloud.
WDR - Wide Dynamic Range, in video surveillance, can impact the quality of images in scenes with strong lightning contrast (e.g., a door opening towards the rising sun). Different techniques, mostly notable multiple exposures, are used to improve WDR image quality. Overall performance has improved over the last decade with key quality differences existing in how well cameras deal with moving objects.
