Video content analysis
Video content analysis (also Video content analytics, VCA) is the capability of automatically analyzing video to detect and determine temporal events not based on a single image. As such, it can be seen as the automated equivalent of the biological visual cortex.
This technical capability is used in a wide range of domains including entertainment,[1] health-care, retail, automotive, transport, home automation, safety and security.[2] The algorithms can be implemented as software on general purpose machines, or as hardware in specialized video processing units.
Many different functionalities can be implemented in VCA. Video Motion Detection is one of the simpler forms where motion is detected with regard to a fixed background scene. More advanced functionalities include video tracking and egomotion estimation.
Based on the internal representation that VCA generates in the machine, it is possible to build other functionalities, such as identification, behavior analysis or other forms of situation awareness.
VCA relies on good input video, so it is often combined with video enhancement technologies such as video denoising, image stabilization, unsharp masking and super-resolution.
Functionalities
Several articles provide an overview of the modules involved in the development of video analytic applications.[3][4] This is a list of known functionalities and a short description.
| Function | Description |
|---|---|
| Dynamic masking | Blocking a part of the video signal based on the signal itself, for example because of privacy concerns. |
| Egomotion estimation | Egomotion estimation is used to determine the location of a camera by analyzing its output signal. |
| Motion detection | Motion detection is used to determine the presence of relevant motion in the observed scene. |
| Shape recognition | Shape recognition is used to recognize shapes in the input video, for example circles or squares. This functionality it typically used in more advanced functionalities such as object detection. |
| Object detection | Object detection is used to determine the presence of a type of object or entity, for example a person or car. Other examples include fire and smoke detection. |
| Recognition | Face recognition and Automatic Number Plate Recognition are used to recognize, and therefore possibly identify persons or cars. |
| Style detection | Style detection is used in settings where the video signal has been produced, for example for television broadcast. Style detection detects the style of the production process.[5] |
| Tamper detection | Tamper detection is used to determine whether the camera or output signal is tampered with. |
| Video tracking | Video tracking is used to determine the location of persons or objects in the video signal, possibly with regard to an external reference grid. |
Commercial applications
VCA is a new technology. New applications are frequently found, however the track record of different types of VCA differs widely. Functionalities such as motion detection and people counting are believed to be available as commercial off-the-shelf products with a decent track-record, even freeware such as dsprobotics Flowstone can handle movement and color analysis.
In many domains VCA is implemented on CCTV systems, either distributed on the cameras (at-the-edge) or centralized on dedicated processing systems. Video Analytics and Smart CCTV are commercial terms for VCA in the security domain. In the UK the BSIA has developed an introduction guide for VCA in the security domain.[6] In addition to video analytics and to complement it audio analytics can also be used.[7]
Kinect is an add-on peripheral for the Xbox 360 gaming console that uses VCA for part of the user input.[8]
The quality of VCA in the commercial setting is difficult to determine. It depends on many variables such as use case, implementation, system configuration and computing platform. Typical methods to get an objective idea of the quality in commercial settings include independent benchmarking[9] and designated test-locations.
Academic research
Video content analysis is a subset of computer vision and thereby of artificial intelligence. Two major academic benchmark initiatives are TRECVID,[10] which uses a small portion of i-LIDS video footage, and the PETS Benchmark Data.[11] They focus on functionalities such as tracking, left luggage detection and virtual fencing. Significant academic research into the field is ongoing at the LIVS, University of Calgary, University of Waterloo, University of Kingston, Georgia Institute of Technology, Carnegie Mellon University, West Virginia University, and The British Columbia Institute of Technology.
References
- ^ KINECT, add-on peripheral for the Xbox 360 console
- ^ VCA usage increase in British Security, BSIA report
- ^ Nik Gagvani, Introduction to Video Analytics
- ^ Cheng Peng, Video Analytics
- ^ Style detection, Cees G.M. Snoek et al., Detection of TV news monologues by style analysis, ICME'04
- ^ British Industry VCA Guide, 262 An Introduction to Video Content Analysis Industry Guide
- ^ UK based startup that provides audio analytics into the CCTV industry
- ^ "Project Natal 101". Microsoft. 2009-06-01. Archived from the original on 2009-06-01. Retrieved 2009-06-02.
- ^ i-Lids, Benchmarking initiative by the UK Home Office
- ^ TRECVID, Academic benchmark initiative by NIST
- ^ PETS Benchmark Data, Performance Evaluation of Tracking and Surveillance (PETS) by University of Reading