Computerized image processing
Description/objectives
Computerised image processing refers to the manipulation of recorded video using computer based software. Once the recorded video has been downloaded onto the PC hard drive it can be accessed using this software platform which provides an interface between the user and the video image processor which performs the manipulation functions. This allows the operator to fast-forward and rewind video at different rates, perform zoom functions, watch video in slow motion, or view video on a frame by frame basis. These are the most basic analytical functions available, but there are more complex functional capabilities available depending on the software being used and the equipment integrated with the surveillance cameras themselves. An example of this is software which allows for video footage to be manipulated in line with vehicle tracking information. This software can be applied to camera systems which are integrated with in-vehicle AVL systems (or have embedded GPS devices). Vehicle movements can be mapped and specific route sections can be chosen by the operator. Once the section is chosen, the software will show which vehicles have passed through at what times of the day, in addition to the locations of on-street cameras in the area. A time frame can be selected, such as the time of an accident, and video footage from each source can be displayed on screen individually or in unison. This allows the operator to view an event from multiple points and to overlap footage and these features could be very useful in verifying liability claims.
These are examples of after-the-fact image processing but processing can also be done in real-time. Examples of these real-time applications include camera based automatic passenger counting (APC) systems or congestion warning systems. APC systems can detect the number of passengers entering and exiting a vehicle, accurately track their trajectory, and distinguish between human and non-human objects such as luggage. Similarly these systems can be used to detect traffic congestion by tracking vehicle numbers, speed and trajectory. This functional capability represents a relatively new aspect of this technology and can potentially be very expensive to implement. These functionalities can however be very useful from an operations management perspective, allowing for route condition monitoring, computer aided dispatch and demand responsive transport, which may justify investment.
Applications
- Operations management
- Demand responsive transport
- Surveillance
Benefits and cautions
While there is still some debate regarding the effectiveness of surveillance systems as a deterrent of criminal activity, it is likely to be better than no system at all. It may also provide peace of mind to transport users making them feel more comfortable on services, particularly late at night. It also has the substantial benefit of being able to provide video evidence of events which could be used in court once video quality is adequate. Computer image processing can be a vital tool in these situations providing a means of accurately identifying what occurred in a given situation.
Relevant case studies
Not observed in the Case Studies
Relevant examples in London (Underground) and Paris (Metro); Singapore