IP Camera & Web Cam

Description

History

The first centralized IP camera, the AXIS Neteye 200, was released in 1996 by Axis Communications.^[3] Although the product was advertised to be accessible from anywhere with an internet connection,^[4] the camera was not capable of streaming real-time video, and was limited to returning a single image for each request in the Common Intermediate Format (CIF). This limitation can be attributed to the lack of powerful integrated circuits capable of handling image processing at the time of release. As a result, the camera was aimed primarily at the tourism industry, and not intended to replace traditional analog CCTV systems.^[5][6]

The first decentralized IP camera was released in 1999 by Mobotix. The camera’s Linux system contained video, alarm, and recording management functions. In 2005, the first IP camera with onboard video content analytics (VCA) was released by Intellio. This camera was able to detect a number of different events, such as if an object was stolen, a human crossed a line, a human entered a predefined zone, or if a car moved in the wrong direction.^[7]

With advancements in cloud infrastructure, Ring (owned by Amazon, U.S.), released its first IP Camera doorbell targeted for home use in 2014. The device offered quick setup, cloud-based recording, and motion detection. The device retailed for $199 USD. As of 2021, Ring has sold millions of units. With the success of IP Cameras, other companies such as Nest (owned by Google, U.S.) released similar cloud-based devices.^[8]

Standard

Previous generations of analog CCTV cameras use established broadcast television formats (e.g. CIF, NTSC, PAL, and SECAM). Since 2000, there has been a shift in the consumer TV business towards high-definition (HD) resolutions (e.g. 1080P (Full-HD), 4K resolution (Ultra-HD) and 16:9 widescreen format).^[9]

IP cameras may differ from one another in resolution, features, video encoding schemes, available network protocols, and the API for video management software.^[10]

To address IP video surveillance standardization issues, two industry groups formed in 2008: the Open Network Video Interface Forum (ONVIF) and the Physical Security Interoperability Alliance (PSIA). PSIA was founded by 20 member companies including Honeywell, GE Security, and Cisco. ONVIF was founded by Axis Communications, Bosch and Sony.^[11] Each group now has numerous additional members, thus cameras and recording hardware that operate under the same standard are compatible with each other.^[10]

Technology

Network Cameras are developed for both enterprise and consumer use. Consumer IP cameras used for home security typically send live video to a companion app on a user’s device. They generally connect to the internet through Wi-Fi or an Ethernet cable.^[12] Unlike consumer IP Cameras, Enterprise IP Cameras often offer higher video resolution, video analytics, and are mostly accessed though HTTP and real time streaming protocol (RTSP).

IP cameras used to be more common in businesses rather than in homes, but that is no longer the case. A 2016 survey of 2,000 Americans revealed 20% of them owned home security cameras. This crossover to IP cameras in home use is partly due to the device’s self-installation. IP cameras typically don’t require professional installation saving time for home and business owners. ^[13]

One of the most popular abilities that consumer-level home security cameras have is to view their footage via a mobile app. Many cameras offer features such as a wide-angle lens, low-light or night vision capabilities, and motion detection. Most are developed to send out notifications via an application such as when motion is detected. Video clips can be stored in a local device such as a micro-SD Card or through a cloud service.^[14]

The market size of home security systems reached $4.8 billion in 2018. It had a compound annual growth rate of 22.4% between 2011 and 2018.^[15] People in countries that suffer from high crime rates, particularly robbery and theft, are keen to adopt home security cameras. The US and China have a high implementation rate of residential security cameras.^[16]

Major key players in the home security market are Nest (owned by Google, U.S.), Ring (owned by Amazon, U.S.), and Arlo (owned by Netgear, U.S.). In the alarm security industry key players are ADT (U.S.), Vivint (U.S.), and SimpliSafe (U.S.). The largest IP Camera manufactures are Hikvision Digital Technology (China), Axis Communications (Sweden), and Dahua (China)^[17]

IP camera types

Depending on their functionality, IP Cameras are generally classified as fixed, varifocal, or pan–tilt–zoom (PTZ camera). Fixed cameras feature an immobile perspective on the subject, whereas varifocal cameras have the ability to remotely adjust the zoom of the image. In addition, PTZ cameras have the ability to direct the camera assembly in any direction remotely. This can be used to track motion or manually adjust the monitoring area. IP Cameras can be designed for indoor or outdoor use. Outdoor cameras are often rated IP65/IP67 in order to withstand outdoor conditions.

IP Cameras can offer a variety of digital imaging technologies such as multi-sensor cameras, panoramic cameras, and thermal imaging cameras.^[18]

Cloud and local storage

Some camera manufacturers offer cloud subscriptions where users may remotely view and download recent video clips by paying recurring subscription fees. Cloud subscription plans typically come with several days of looping storage, and the videos will be overwritten beyond this duration.^[19]

Some cameras include a micro SD card slot so users can store videos locally. Most IP Cameras can be programmed to overwrite old video once the storage medium is full. Accessing the video on the camera can normally be done via a direct network connection to the device.

Considerations

Potential benefits

Previous generation cameras transmitted analog video signals. IP cameras send images digitally using the transmission and security features of the TCP/IP protocol. Advantages to this approach include:

• Two-way audio via a single network cable allows users to listen to and speak to the subject of the video (e.g., a clerk assisting a customer through step-by-step instructions)
• Use of a Wi-Fi or wireless network^[20]
• Distributed artificial intelligence (DAI)—as the camera can contain video analytics that analyze images^[21]
• Secure data transmission through encryption and authentication methods such as WPA or WPA2, TKIP or AES
• Remote accessibility that lets users view live video from any device with sufficient access privileges^[22]
• Power over Ethernet (PoE) to supply power through the Ethernet cable and operate without a dedicated power supply
• Better image resolution, typically four times the resolution of an analog camera^[23]

Artificial intelligence and Internet privacy

The American Civil Liberties Union (ACLU) has expressed privacy concerns if AI is widely practiced. AI is capable of tracking movements and studying behaviors; moreover, AI can also recognize emotions, and further predict patterns of movement.^[24]

Facial recognition system

Facial recognition identifies a human face by analyzing facial features from a picture or video, an example of biometrics. If a camera allows users to set up a database that includes family members and close friends, the system may distinguish whether someone exists in the database. If the camera is capable of providing accurate facial recognition, it can tell if the person it detects is authorized (in the database). The detection of unauthorized persons may prompt the owner to call law enforcement.^[25] The footage can be used as a means of identifying and apprehending offenders.

Potential concerns

Concerns include:

• Privacy concerns^[26]
• Average higher purchase cost per camera^[27]
• Security can be compromised by insecure credentials, given that the camera can be accessed independently of a video recorder.
• Public internet connection video can be complicated to set up^[28] or using the peer-to-peer (P2P) network.
• Data storage capacity concerns^[29]

Hacking

If video is transmitted over the public internet rather than a private network or intranet, CCTV devices potentially becomes open to a wider audience including hackers. Malicious actors can access private CCTV systems to disable, manipulate, or observe security measures– as well as gain further access to the private network it’s connected to, often referred to colloquially as pivoting. This risk can be mitigated by securing the network via Firewall rules designed to restrict access to the device, and by keeping software and firmware up to date.

In 2012, users of 4chan hacked into thousands of streaming personal IP cameras by exploiting a vulnerability in some models of TRENDnet home security cameras.^[30] In 2014, it was reported that a site indexed 73,011 locations worldwide with security cameras that used default usernames and passwords, and were therefore, unprotected.^[31] Automated services like Shodan.io constantly scan residential and commercial IP blocks to automatically detect and catalog open ports and services, including those commonly used for IP cameras.^[32]

History

Early development (early 1990s)

First developed in 1991, a webcam was pointed at the Trojan Room coffee pot in the Cambridge University Computer Science Department (initially operating over a local network instead of the web). The camera was finally switched off on August 22, 2001. The final image captured by the camera can still be viewed at its homepage.^[1][2] The oldest continuously operating webcam, San Francisco State University‘s FogCam, has run since 1994 and is still operating as of October 2022. It updates every 20 seconds.^[3]

The released in 1993 SGI Indy is the first commercial computer to have a standard video camera,^[4] and the first SGI computer to have standard video inputs.^[5]

The maximum supported input resolution is 640×480 for NTSC or 768×576 for PAL. A fast machine is required to capture at either of these resolutions, though; an Indy with slower R4600PC CPU, for example, may require the input resolution to be reduced before storage or processing. However, the Vino hardware is capable of DMAing video fields directly into the framebuffer with minimal CPU overhead.

The first widespread commercial webcam, the black-and-white QuickCam, entered the marketplace in 1994, created by the U.S. computer company Connectix. QuickCam was available in August 1994 for the Apple Macintosh, connecting via a serial port, at a cost of $100. Jon Garber, the designer of the device, had wanted to call it the “Mac-camera”, but was overruled by Connectix’s marketing department; a version with a PC-compatible parallel port and software for Microsoft Windows was launched in October 1995. The original QuickCam provided 320×240-pixel resolution with a grayscale depth of 16 shades at 60 frames per second, or 256 shades at 15 frames per second.^[6] These cam were tested on several Delta II launch using a variety of communication protocols including CDMA, TDMA, GSM and HF.

Videoconferencing via computers already existed, and at the time client-server based videoconferencing software such as CU-SeeMe had started to become popular.

The first widely known laptop with integrated webcam option, at a pricepoint starting at US$ 12,000, was an IBM RS/6000 860 laptop^[7][8] and his ThinkPad 850 sibling,^[9] released in 1996.

Entering the mainstream (late 1990s)

One of the most widely reported-on webcam sites was JenniCam, created in 1996, which allowed Internet users to observe the life of its namesake constantly, in the same vein as the reality TV series Big Brother, launched four years later.^[10] Other cameras are mounted overlooking bridges, public squares, and other public places, their output made available on a public web page in accordance with the original concept of a “webcam”. Aggregator websites have also been created, providing thousands of live video streams or up-to-date still pictures, allowing users to find live video streams based on location or other criteria.

In the late 1990s, Microsoft NetMeeting was the only videoconferencing software on PC in widespread use, making use of webcams.^[11] In the following years, instant messaging clients started adding webcam support: Yahoo Messenger introduced this with version 5.5 in 2002, allowing video calling in 20 frames per second using a webcam.^[12] MSN Messenger gained this in version 5.0 in 2003.^[13]

2000s–present

Around the turn of the 21st century, computer hardware manufacturers began building webcams directly into laptop and desktop screens, thus eliminating the need to use an external USB or FireWire camera. Gradually webcams came to be used more for telecommunications, or videotelephony, between two people, or among several people, than for offering a view on a Web page to an unknown public.

For less than US$100 in 2012, a three-dimensional space webcam became available, producing videos and photos in 3D anaglyph image with a resolution up to 1280 × 480 pixels. Both sender and receiver of the images must use 3D glasses to see the effect of three dimensional image.^[14]

Webcams are considered an essential accessory for remote work, mainly to compensate for lower quality video processing with the built-in camera of the average laptop. During the COVID-19 pandemic, there was a shortage of webcams.^[15] Most laptops before and during the pandemic were made with cameras capping out at 720p recording quality at best, compared to the industry standard of 1080p or 4K seen in smartphones and televisions from the same period.^[16] The backlog on new developments for built-in webcams is the result of a design flaw with laptops being too thin to support the 7mm camera modules to fit inside, instead resorting to ~2.5mm.^[17][18] Also the camera components are more expensive and not a high level of demand for this feature, companies like Apple have not updated their webcams since 2012.^[19] Smartphones started to be used as a backup option or webcam replacement, with kits including lighting and tripods or downloadable apps.^[20]

Technology

Image sensor

Image sensors can be CMOS or CCD, the former being dominant for low-cost cameras, but CCD cameras do not necessarily outperform CMOS-based cameras in the low-price range. Most consumer webcams are capable of providing VGA-resolution video at a frame rate of 30 frames per second. Many newer devices can produce video in multi-megapixel resolutions, and a few can run at high frame rates such as the PlayStation Eye, which can produce 320×240 video at 120 frames per second.^[21]

As webcams evolved simultaneously with display technologies, USB interface speeds and broadband internet speeds, the resolution went up from gradually from 320×240, to 640×480, and some now even offer 1280×720 (aka 720p) or 1920×1080 (aka 1080p) resolution.^[22][23][24] Despite the low cost, the resolution offered as of 2019 is impressive, with now the low-end webcams offering resolutions of 720p, mid-range webcams offering 1080p resolution, and high-end webcams offering 4K resolution at 60 fps.

Optics

Various lenses are available, the most common in consumer-grade webcams being a plastic lens that can be manually moved in and out to focus the camera. Fixed-focus lenses, which have no provision for adjustment, are also available. As a camera system’s depth of field is greater for small image formats and is greater for lenses with a large f-number (small aperture), the systems used in webcams have a sufficiently large depth of field that the use of a fixed-focus lens does not impact image sharpness to a great extent.

Most models use simple, focal-free optics (fixed focus, factory-set for the usual distance from the monitor to which it is fastened to the user) or manual focus.

Webcams can come with different presets and fields of view. Individual users can make use of less than 90° horizontal FOV for home offices and live streaming. Webcams with as much as 360° horizontal FOV can be used for small- to medium-sized rooms (sometimes even large rooms). Depending on the users’ purposes, webcams in the market can display the whole room or just the general vicinity.

Internal software

As the bayer filter is proprietary, any webcam contains some built-in image processing, separate from compression. Digital video streams are represented by huge amounts of data, burdening its transmission (from the image sensor, where the data is continuously created) and storage alike. Most if not all cheap webcams come with built-in ASIC to do video compression in real-time.

Support electronics read the image from the sensor and transmit it to the host computer. The camera pictured to the right, for example, uses a Sonix SN9C101 to transmit its image over USB. Typically, each frame is transmitted uncompressed in RGB or YUV or compressed as JPEG. Some cameras, such as mobile-phone cameras, use a CMOS sensor with supporting electronics “on die”, i.e. the sensor and the support electronics are built on a single silicon chip to save space and manufacturing costs. Most webcams feature built-in microphones to make video calling and videoconferencing more convenient.

Interface and external software

Typical interfaces used by articles marketed as a “webcam” are USB, Ethernet and IEEE 802.11 (denominated as IP camera). Further interfaces such as e.g. Composite video, S-Video or FireWire were also available. The USB video device class (UVC) specification allows inter-connectivity of webcams to computers without the need for proprietary device drivers.

Various proprietary as well as free and open-source software is available to handle the UVC stream. One could use Guvcview or GStreamer and GStreamer-based software to handle the UVC stream. Another could use multiple USB cameras attached to the host computer the software resides on, and broadcast multiple streams at once over (Wireless) Ethernet, such as MotionEye. MotionEye can either be installed onto a Raspberry Pi as MotionEyeOs, or afterwards on Raspbian as well. MotionEye can also be set up on Debian, Raspbian is a variant of Debian. Note that MotionEye V4.1.1 ( Aug ’21 ) can only run on Debian 10 Buster ( oldstable ) and Python 2.7. Newer versions such as 3.X are not supported at this point of time according to Ccrisan, foundator and author of MotionEye.

Various software tools in wide use can be employed to take video and pictures, such as PicMaster and Microsoft’s Camera app (for use with Windows operating systems), Photo Booth (Mac), or Cheese (with Unix systems). For a more complete list see Comparison of webcam software.

Uses

The most popular use of webcams is the establishment of video links, permitting computers to act as videophones or videoconference stations. For example, Apple‘s iSight camera, which is built into Apple laptops, iMacs and a majority of iPhones, can be used for video chat sessions, using the Messages instant messaging program. Other popular uses include security surveillance, computer vision, video broadcasting, and for recording social videos.

Videotelephony

Webcams can be added to instant messaging, text chat services such as AOL Instant Messenger, and VoIP services such as Skype, one-to-one live video communication over the Internet has now reached millions of mainstream PC users worldwide. Improved video quality has helped webcams encroach on traditional video conferencing systems. New features such as automatic lighting controls, real-time enhancements (retouching, wrinkle smoothing and vertical stretch), automatic face tracking and autofocus, assist users by providing substantial ease-of-use, further increasing the popularity of webcams.

Webcams can also encourage remote work, enabling people to work remotely via the Internet. This usage was crucial to the survival of many businesses during the COVID-19 pandemic, when in-person office work was discouraged. Businesses, schools, and individuals have relied on video conferencing instead of spending on business travel for meetings. Moreover, the number of video conferencing cameras and software have multiplied since then due to their popularity.

Webcam features and performance can vary by program, computer operating system, and also by the computer’s processor capabilities. Video calling support has also been added to several popular instant messaging programs.

Webcams allow for inexpensive, real-time video chat and webcasting, in both amateur and professional pursuits. They are frequently used in online dating and for online personal services offered mainly by women when camgirling. However, the ease of webcam use through the Internet for video chat has also caused issues. For example, moderation system of various video chat websites such as Omegle has been criticized as being ineffective, with sexual content still rampant.^[25] In a 2013 case, the transmission of nude photos and videos via Omegle from a teenage girl to a schoolteacher resulted in a child pornography charge.^[26]

The popularity of webcams among teenagers with Internet access has raised concern about the use of webcams for cyber-bullying.^[27] Webcam recordings of teenagers, including underage teenagers, are frequently posted on popular Web forums and imageboards such as 4chan.^[28][29]

Monitoring

Webcams can be used as security cameras. Software is available to allow PC-connected cameras to watch for movement and sound, recording both when they are detected. These recordings can then be saved to the computer, e-mailed, or uploaded to the Internet. In one well-publicised case,^[30] a computer e-mailed images of the burglar during the theft of the computer, enabling the owner to give police a clear picture of the burglar’s face even after the computer had been stolen.

In December 2011, Russia announced that 290,000 Webcams would be installed in 90,000 polling stations to monitor the 2012 Russian presidential election.^[31] Webcams may be installed at places such as childcare centres, offices, shops and private areas to monitor security and general activity.

Astrophotography

With very-low-light capability, a few specific models of webcams are very popular to photograph the night sky by astronomers and astro photographers. Mostly, these are manual-focus cameras and contain an old CCD array instead of comparatively newer CMOS array. The lenses of the cameras are removed and then these are attached to telescopes to record images, video, still, or both. In newer techniques, videos of very faint objects are taken for a couple of seconds and then all the frames of the video are “stacked” together to obtain a still image of respectable contrast.^[32]

Laser beam profiling

A webcam’s CCD response is linearly proportional to the incoming light.^[33] Therefore, webcams are suitable to record laser beam profiles, after the lens is removed. The resolution of a laser beam profiler depends on the pixel size. Commercial webcams are usually designed to record color images. The size of a webcam’s color pixel depends on the model and may lie in the range of 5 to 10 µm. However, a color pixel consists of four black and white pixels each equipped with a color filter (for details see Bayer filter). Although these color filters work well in the visible, they may be rather transparent in the near infrared. By switching a webcam into the Bayer-mode it is possible to access the information of the single pixels and a resolution below 3 µm was possible.^[34]

Privacy concerns

Many users do not wish the continuous exposure for which webcams were originally intended, but rather prefer privacy.^[35] Such privacy is lost when malware allow malicious hackers to activate the webcam without the user’s knowledge, providing the hackers with a live video and audio feed.^[36] This is a particular concern on many laptop computers, as such cameras normally cannot be physically disabled if hijacked by such a Trojan Horse program or other similar spyware programs.

Cameras such as Apple‘s older external iSight cameras include lens covers to thwart this. Some webcams have built-in hardwired LED indicators that light up whenever the camera is active, sometimes only in video mode.^[37] However, it is possible for malware to circumvent the indicator and activate the camera surreptitiously, as researchers demonstrated in the case of a MacBook‘s built-in camera in 2013.^[37]

Various companies sell sliding lens covers and stickers that allow users to retrofit a computer or smartphone to close access to the camera lens as needed.^[37] One such company reported having sold more than 250,000 such items from 2013 to 2016.^[37] However, any opaque material will work just as well.^[37]

The process of attempting to hack into a person’s webcam and activate it without the webcam owner’s permission has been called camfecting, a portmanteau of cam and infecting. The remotely activated webcam can be used to watch anything within the webcam’s field of vision. Camfecting is most often carried out by infecting the victim’s computer with a computer virus.