GB2420636A

GB2420636A - Road transport network incident or accident re-construction method

Info

Publication number: GB2420636A
Application number: GB0422331A
Authority: GB
Inventors: David Iwan Parry
Original assignee: TRL Ltd
Current assignee: TRL Ltd
Priority date: 2004-10-08
Filing date: 2004-10-08
Publication date: 2006-05-31
Also published as: GB0422331D0

Abstract

A method of re-constructing a road transport network incident or accident comprises the step of obtaining image data, using an array of camera units 150-158, prior to and/or during the incident. The method comprises the steps of determining from the image data whether sufficient information has been obtained to re-construct the road transport network incident or accident; and basing an operational road transport network decision on an outcome of the determining step. The determination step may be performed automatically. In this manner, the time taken to re-open a road transport network after a network incident can be minimised. This aids the reduction of congestion on the network.

Description

- 1 - 2420636

ROAD TRANSPORT NETWORK INCIDENT RE-CONSTRUCTION SYSTEM

AND METHOD THEREFOR

Field of the Invention

This invention relates to an improved mechanism for re- constructing a road transport network incident. The invention is applicable to, but not limited to, a system comprising an array of cameras to identify movements of vehicles prior to an incident using camera images.

Background of the Invention

With the ever-increasing use of vehicles, congestion on roads is becoming problematic. Much congestion is due to road accidents on major roads, such as trunk roads and:.,.

motorways. A reduction in the time taken to clear the scenes of traffic incidents on the trunk road and * S *** SS S motorway network could make a significant contribution to one of the UK Highways Agency's key objectives, namely: "Making journeys more reliable through better *S..

network management and information" : S...

In the UK and internationally, Police and civilian Investigators are required to identify, gather and map detailed evidence at all road traffic collision scenes that involve either fatality or life-threatening injury.

This work is central to conducting effective investigation and (in the UK) to meeting the requirements of the Association of Chief Police Officer's (ACPO) Road Death Investigation Manual (RDIM). The RDIM requires that all road traffic incidents that involve fatality or life threatening injury be investigated as either an unlawful killing or a potential unlawful killing.

Such evidence must be gathered in as safe and controlled a manner as possible. Currently, evidence is gathered through on-road investigation. These investigations require the detailed inspection of accident marks, such as tyre marks, scrapes, gouges etc. Such activities require lane or carriageway closures. Importantly, the identification, measuring and mapping of crucial physical evidence at incident scenes, by investigators using existing methods, is very time consuming. This time is almost always a contributory factor to the delays associated with the ultimate clear up of scenes and the re-opening of the road. * * S * **

It should be recognised, however, that road user delays: associated with road traffic incidents are also * S S..

influenced by a number of additional factors, which are S.....

independent of the time currently taken by investigators to measure physical evidence at the incident scene. Such, factors can include: : S...

(i) The recovery of disabled vehicles from traffic lanes; (ii) The repair of damaged infrastructure; (iii) The clearance of debris and/or the disbursement of spillage; and (iv) The time taken to extricate trapped persons in damaged vehicles.

One of the major issues behind the identification of alternative methods of scene data collection is to counter the potential for road traffic incident scene investigation times to increase. A focus of such work, resulting from the RDIM is the identification of equipment capable of measuring required evidence at such scenes in a much shorter time.

With the increasing pressure on road operators and police services to manage incidents more holistically and effectively, it is important that more time efficient techniques and/or practices are developed.

Currently, police officers use surveying systems at traffic accident scenes to record crucial physical evidence. These systems are termed total stations'. A total station is a surveying tool that is used to measure three-dimensional points precisely. The recording of crucial evidence in this way may take significantly: :s longer than the time taken to clear the physical debris S.....

from the scene.

20:...:.

Surveying systems map and record the physical evidence S...

left behind by the incident. Notably, the only S...

information used in this regard, is the information obtained directly from the scene itself. Investigating officers use this information after the crash. These officers reconstruct' the speeds and relative movements of the vehicles involved in the crash. The reconstruction can cover speeds and relative movements both at the time of the crash and in the moments immediately preceding it. However, as mentioned, such approaches for determining vehicle movements that lead up to a traffic incident are extremely slow, and generally cause major disruptions to the trunk road and highway networks. The reconstruction may be performed whilst the carriageway or road is still closed. If this is the case, this may add to the time taken merely to record physical evidence at the scene, such as measuring skid and gauge marks. However, the primary aim is typically to record the physical evidence that will allow the reconstruction to be performed in the office'.

A related field of technology to crash investigation is the field of crash detection'. In most road networks today, drivers, passengers or pedestrians witness an accident. They then assist at the scene, and/or summon emergency services. However, systems have been developed in Japan that can detect automatically the occurrence of a road crash, or near miss. * .* * S * a..

One known system for automated crash detection', or a....

detection of a network incident, comprises a series of as....

microphones. The microphones are deployed at a particular location, such as a road junction. The *5** microphones are able to recognise the occurrence of a **..

crash or near-miss' by detecting particular sounds.

These might be the noise frequencies associated with the sound of brakes or tyres under heavy braking, or the sound of vehicle impacts. Such a detection system may be linked to cameras. The detection system may trigger the camera system to record an accident or near-miss' as it occurs at the particular location under surveillance.

This system is predominantly used to help analyse why accidents occur at a particular location.

Thus, there is great pressure to reduce the time currently required for measurement at an accident scene, and for the reconstruction of events by expert investigation officers. A solution to this problem would substantially alleviate the abovementioned disadvantages.

Statement of Invention

In accordance with a first aspect of the present invention there is provided a method of reconstructing a road transport incident, as claimed in Claim 1.

In accordance with a second aspect of the present invention there is provided a signal processing unit, as claimed in Claim 32. * * * * S.

In accordance with a third aspect of the present invention there is provided an array of cameras, as S.....

claimed in Claim 33.

In accordance with a fourth aspect of the present s.. *.

invention there is provided a road transport network S...

incident re-construction system, as claimed in Claim 16.

Thus, in summary, the aforementioned problems with known techniques are resolved by using a series of cameras recording images and locations of traffic prior to an incident. A determination from a series of still camera images can then be made quickly to ascertain whether the information is of sufficient detail and reliability to accurately reconstruct the actions of vehicles up to the incident. In this manner, it is possible to ascertain a attempting to relate these tyre marks to potentially one of a number of prone vehicles at an incident.

The proposed system of (crash' or traffic safety') cameras removes the need for detailed on-scene investigations by police officers, thereby allowing recovery crews to facilitate a significantly earlier opening of the road to traffic. The crash' or traffic safety' camera systems preferably operate by constantly recording traffic movements through long arrays of substantially regularly spaced cameras with fixed views of the road carriageway. During an incident these cameras (which are constantly recording images of the road) record the movements of the vehicles involved in the crash in the moments before and during the incident.

This allows: (i) The relative movements of vehicles to be observed and quantified by video/camera analysis; and (ii) The speeds of the vehicles to be calculated.

Brief Description of the Drawings

Exemplary embodiments of the present invention will now be described, with reference to the accompanying drawings, in which: FIG. 1 illustrates an overview of a camera system used in accordance with a preferred embodiment of the invention; and FIG. 2 is a flowchart illustrating a method of reconstructing a road transport network incident by detecting vehicle movements that lead up to a traffic incident, in accordance with a preferred embodiment of the invention.

Description of Preferred Enthodiments

The most common form of 2D image generation is a picture that is taken by a camera. Camera units are actively used in many environments. In some instances, where pictures are required to be taken from a number of locations, multiple camera units may be used and an Operator may view the pictures remotely.

In the context of the present invention, an array of camera units is configured to be operably linked to an image-processing unit that processes images captured by one or more camera units. An example of an array of camera units that is suitable for such a purpose would consist of standard or enhanced resolution colour or monochrome all weather cameras (or cameras in weather proof housing), which have reliable performance in a wide range of temperature conditions, and would preferably be installed within suitable protective enclosures. Many manufacturers, such as Pelcom or Vid0TN, produce camera types that could meet the requirements of the crash camera systems described herein.

Referring now to FIG. 1, a road traffic incident detection system 100 is illustrated. The road traffic incident detection system 100 shows a highway 110, with vehicles 130 moving along the highway. A number of camera units 150, 152, 154, 156 and 158 are located preferably at substantially regular spaced intervals 120, 122, 124, 126, and 128 above the highway 110. The field of view of the camera units are preferably arranged to overlap. In this manner, a vehicle identified by a captured image by a first camera unit, is also captured by one or more further camera units along the highway 100. As such, the movement of the vehicle(s) along the highway 100 are always captured in an image by one or more camera unit(s), with no gaps in the highway images.

S.....

The series/array of camera units is preferably directed in substantially the same direction and focused on the direction of traffic flow along the highway. However, it is envisaged that the configuration/location of camera units may comprise two or more camera units, which are directed in opposite directions, and/or directed to either side of the highway. In such a manner, one camera unit may capture images from both sides of the highway, i.e. capturing images of vehicles moving in opposite directions.

Alternatively, instead of being located directly over the highway, there may be some instances, for example to track accident black spots' on bends, that camera units may be positioned on the side of the road, and may be spaced at irregular intervals, for example to better capture the movement of vehicles around a bend in the road.

Notably, the camera units are operably coupled to a signal processing unit 160, which is arranged to process image data sent thereto. It is envisaged that the image data may be sent 140, 142, 144, 146, and 148 via either wireline or wireless means. It is also envisaged that the signal processing unit 160 may be part of a central signal processing system, or processing software in a computer. The signal processing unit 160 may also be located remotely from the camera units and arranged to specifically analyse captured image data from a number of..

highways, as and when requested or required by an appropriate Authority such as the Police Services.

It is envisaged that by collating image data from a plurality of camera units, it is possible that the S...

vehicular motion of a number of vehicles may be determined, i.e. reconstructed. In this context, the term vehicular motion' comprises, at least, direction(s) that a vehicle takes during the lead up to an incident, and/or speed(s) that the vehicle was moving.

Processing of the sequence of images recorded by each camera within the array allows individual moving blocks' (vehicles), within the set of images produced by each camera, to be identified and monitored by processing software. The processing software may provide an alert - 10 relating to the detection (or possible detection) of an incident. This would be where certain criteria are satisfied within the processing software relating to the nature of vehicle movements.

All images produced by all cameras are preferably uniquely referenced with respect to the time of the image (time/date stamping) and the source of the image (camera reference number) . This timestamp information, coupled with precise knowledge as to the manner in which the camera system captures images specifically the precise frame rate and shutter speed timing, is preferably used to calculate the speed of involved vehicles immediately before and during road traffic incidents witnessed by the cameras. S. S * . . * SS

It is envisaged that each camera unit will take S..

photographic images (frames) on a regular basis. For S.....

example, the locations of successive cameras may be such that each camera captures five frames (photographic images) of a particular vehicle travelling at normal S...

speeds to identify its movements, before a subsequent S...

camera unit takes subsequent photographs. It is envisaged that a suitable image capture rate would be ten frames per second. That way, the taking of subsequent images can identify individual vehicles, and identify the vehicle's movements that lead up to where each vehicle come to rest after the traffic incident.

Advantageously, relatively poor image data and low- quality camera units can be used, as it is not necessary to identify a vehicle by its number plate, which would - 11 - require many pixels per image. It is only necessary to distinguish the movements of particular vehicles that lead up to where the vehicle rested. Thus, the amount of image data can be limited, as only the tracking of vehicle shapes is needed, with the vehicle's movements re-constructed backwards from its known resting position with the use of the captured images.

Notably, the proposed concepts are not solely a mechanism to identify speed and/or erratic vehicle movements, as unusual events may also be recorded, for example a lorry losing its load, or blowing over in high winds. Notably, the proposed technique preserves information relating to the incident itself, rather than performing a re- construction of the incident, per se. * * . * **

Specifically, by preserving images of the incident: *.

sequence in such a way that the characteristics of the camera system (frame timing, field of view, etc) are known, the speed and relative movements of vehicles seen moving in the images produced by any one camera can be S..

determined. In recording images of the incident sequence the cameras preserve' the incident for later analysis.

This is in contrast to the case of on-scene' police investigations where evidence is collected after an accident to determine the pre/post accident movements of the involved vehicles in order to, where possible, reconstruct' what has occurred only at that scene. The uncertainty involved in the latter is much greater than that which would occur given the use of the system of the preferred embodiment of the present invention as described herein.

- 12 - It is envisaged that the inventive concept of the present invention can be applied to one or more camera units that may be fixed or moveable throughout a range of horizontal and/or vertical directions. In this context, an array of camera units may be located substantially perpendicular to a traffic flow direction on a highway, at a particular location. A similar array of further camera units is preferably located at distal points along the highway itself, and so on. The subsequent array of camera units preferably has an overlap region of captured image data with one or more of the previous array of camera units.

In this manner, multiple camera units can be placed along a highway to monitor traffic flow thereon.

Additionally, it is envisaged that the photographic images obtained by the camera units may be continuously: .".

replaced with one or more updated photographs, to minimise storage of unnecessary data. In this context, image data would have been requested from the camera unit.m.

within, say, an hour of any major traffic incident.

Thus, by employing a one-hour loop of capturing image data, no longer than an hour's worth of image data needs to be stored for a particular camera unit or array of camera units.

Referring now to FIG. 2, a flowchart 200 illustrates a preferred method for detecting and re-constructing a road transport network incident, i.e. determining vehicular motion of traffic prior to a traffic incident. The preferred method comprises a number of camera units, preferably an array of cameras, capturing image data of - 13 - traffic flow on a highway, as shown in step 205. The captured image data may be stored locally, say as a series of bitmaps, as shown in step 215. Alternatively the captured image data may be transmitted directly to a central image data processing system or unit, either via wireline or wireless means, as in step 210.

It is envisaged that image data would be continually captured by all or part of the crash camera array during periods of operation (although the system as a whole is intended to operate continually). Image data would be preserved for a defined time period, whereby, new' data would continually over-write old' data until an incident detection occurred (where the incident was detected either automatically via image processing, or manually) At this point all current data stored by the system would:.

be copied/downloaded to a safe' area to await analysis.

If the captured image data is stored locally, in step 215, the captured image data is then preferably transmitted to a central image data processing system or S...

unit, either via wireline or wireless means, either I...

intermittently (say following a polling operation) or when specifically requested, as shown in step 220.

Following a road transport network incident, it is envisaged that the Police Service is provided with the opportunity to request a series of camera images, over a period of time, from an array of camera units. Thus, the Police are provided with the opportunity to re-locate temporarily stored images in a storage unit that would not be overwritten with new images as they are captured.

- 14 - Alternatively, or in addition, the Police are provided with the opportunity to have specific images downloaded to the central repository for subsequent processing.

Thus, in a further embodiment of the present invention, where multiple camera units are used and the image data information is stored locally, it is envisaged that a polling operation for retrieving captured images from a series of camera units may be employed. Thus, in this context, a central image data processing system would be contacted by, say, the Police when called in response to an incident. The central image data processing system would then poll a particular number and location of camera units on a particular highway, to have the corresponding image data sent to the central image data processing system. The central image data processing system will also include a timestamp to indicate the: .".

timing of captured frames that it wishes to analyse. * S

*5S*S*

S

The image data from a number of camera units is then processed by the central image data processing system, as shown in step 225. The processing of camera image data S...

may result in the images being displayed to an Operator.

The Operator would then determine, as shown in step 230, the reliability of the image data to judge whether it is of sufficient quality in identifying a motion of a number of vehicles leading up to the traffic incident.

Alternatively, it is envisaged that the captured image data may be processed on a bit-wise basis, such that an automatic' determination can be made as to the movements of one or more vehicles, based on bit-by-bit (or image block-by-image block) comparison between successive - 15 captured frames. In this context, an image block could be set to be a particular vehicle, and the movements of that vehicle (substantially a block of bits) tracked automatically.

In this manner, as shown in step 235, the motion of a number of vehicles that lead up to a traffic incident may be re-constructed based on the collated image data from an array of camera units. Furthermore an operational road transport decision can then be made, as shown in step 240, as to whether sufficient image data has been collated to reconstruct the scene' and thereby allow the road to be re-opened as soon as the debris/vehicles have been cleared.

Thus, in an enhanced embodiment of the present invention it is envisaged that automatic detection of changes in: bit/pixel values between captured images, and perhaps a stored clean' image, may be made to ascertain whether the identified changes in vehicular movements (and therefore pixel bits) are sufficient to determine the motion of vehicles that lead up to the road traffic incident. In this context, an image encoder at each camera unit may be configured to only transmit bit/pixel values relating to the change or may transmit the whole image to the central image data processing system. As such, it is envisaged that the image encoder (say, at the camera unit) may not need to transmit any new differential' information if a change, determined between a currently viewed image frame and a stored frame, is below a predetermined threshold.

- 16 - It is envisaged that a threshold process may be used in this regard. It is envisaged that threshold values would relate to the difference in pixel characteristic, whereby no information would be transmitted for certain pixels for successive images where the pixel characteristics are not significantly different to the same pixel in the preceding image. This would serve to reduce the volume of data transmission and storage. It is envisaged that regular full' frames, say one every second, would be required to assist the reconstruction of images stored using this process.

It is envisaged that the inventive concepts described herein can be advantageously utilised in a wide range of applications. For example, it is envisaged that one suitable application would be re-opening roads after a traffic incident. By processing image data in this way, : a determination can be made within a matter of minutes as to whether the captured image data contains sufficient information to identify the motion of all vehicles involved in the incident, prior to the incident pa. .

happening.

Although the preferred embodiment of the present invention is described with reference to an array of camera units located along a direction of traffic flow of a road/highway, it is envisaged that alternative configurations may exist that benefit from the inventive concepts described herein. For example, it is envisaged that the array of camera units' may be a single camera unit capturing multiple images. In this regard, the single camera unit may be focused on a particular traffic - 17 - black-spot', i.e. an area of road network renowned for accidents. Thus, an array of camera units' connected to an image processing unit, should be interpreted as one or more' camera units.

Furthermore, it is envisaged that the image processing may be performed within the camera unit itself, with the processed image transferred to an image repository for subsequent analysis/further processing with a number of other captured images.

In the context of the present invention, it is envisaged that an Operator, or Police Officer investigating the road traffic incident, will be able to determine a variety of facts relating to a number of parameters of a vehicle's motions that lead up to the traffic incident from the captured images. The details of the vehicle motion may comprise: (i) A speed at which a vehicle was travelling.

(ii) The direction that a vehicle was travelling, i.e. whether it was travelling erratically or dangerously, etc. (iii) The influence of other causal factors to the road traffic incident, such as weather effects (flooding, cross carriageway flow of water, wind), or the presence of debris, unauthorised road users, or animals, etc. Where there is a series of camera units located along a highway, the processing unit can then monitor the - 18 - movements of vehicles leading up to the traffic incident, by processing the successive images captured by individual camera units. This monitoring provides one method by which traffic incidents may be detected. It is envisaged that the vehicles moving on the road carriageway will be seen' by the processing unit as blocks' of pixels that move within the successive images generated by an individual camera. It is envisaged that continual monitoring of vehicles by successive cameras could be achieved by the processing unit predicting when a vehicle will become visible to the next camera in the array. Thereafter, the processing unit transfers the monitoring of that vehicle/'block' to the next camera.

Such a feature would not, however, be critical to the basic operation of the proposed system.

The identification of a specific vehicle within any camera image requires some portion of the vehicle to be visible from the camera positioned to monitor the portion....

of the road that the vehicle is travelling on. Full or partial obscuration of a small vehicle may occur with S...

respect to the camera by the passage of the small vehicle...

behind a larger vehicle. In these cases vehicle blocks' will overlap, however the individual portions of the overlapping blocks' could be used for analysis purposes, provided that a sufficient number of pixels of the vehicle block' are visible to distinguish it from other vehicle blocks'.

It is envisaged that the proposed technique is also applicable to both wired and wireless connections/links between one or more distinct camera units or arrays of - 19 - camera units and the remote image data processing unit.

A wireless connection allows the particular benefit of determining vehicle motion of traffic prior to a traffic incident remotely at a dedicated image-processing centre.

It will be understood that the road transport network incident detection system and method of re-constructing a road transport network incident, as described above, aim to provide one or more of the following advantages: (1) A much faster method of determining a motion of a number of vehicles can be made to identify a cause of a traffic incident; (ii) The system of crash' or traffic safety cameras' removes the need for detailed on-scene investigations by police officers, thereby allowing:.

recovery crews to focus on the early opening of the road to traffic; and (iii) It is expected that the presence of crash' i'...

or traffic safety cameras' would passively encourage road safety by providing the authorities with the means S...

of more accurate accident investigation and a method to S...

show the nature of driver's actions that caused or contributed to a road traffic accident.

Thus, in summary, a method of re-constructing a road transport network incident has been described. The method comprises the steps of obtaining image data from an array of camera units. The method further comprises the steps of determining whether sufficient information has been obtained tore-construct the road transport network incident, and basing an operational road transport decision on an outcome of the determining step.

- 20 - Preferably, the road transport network incident may be subsequently re- constructed based on the processed image data.

Furthermore, a road transport network incident re- construction system has been described. The system comprises an array for processing obtained image data to ascertain that a network incident has occurred. In particular, the signal processing unit determines whether sufficient information has been obtained to re-construct the road transport network incident such that the road transport network incident can be re-constructed subsequently, based on the processed image data.

Whilst the specific and preferred implementations of the:* embodiments of the present invention are described above, it is clear that one skilled in the art could readily apply variations and modifications that would still employ the aforementioned inventive concepts.

Thus, a road transport network incident re-construction S...

system comprising an array of camera units and a method of reconstructing a road transport network incident have been provided wherein the abovementioned disadvantages with prior art arrangements have been substantially alleviated.

Claims

- 21 - Claims 1. A method of re-constructing a road transport network

incident (200) comprising the step of: obtaining image data, using an array of camera units (205), prior to and/or during the incident; wherein the method is characterised by the steps of: determining (230) from the image data whether sufficient information has been obtained to re-construct the road transport network incident; and basing (240) an operational road transport decision on an outcome of the determining step (230).
2. A method of re-constructing a road transport network incident (200) according to Claim 1, further:.

characterised by the steps of processing the image data, and reconstructing (235) the road transport network incident based on the processed image data. * *

S.....
3. A method of re-constructing a road transport *5*S network incident (200) according to Claim 1 or Claim 2, S...

wherein the step of basing an operational road transport **.

decision comprises the step of clearing away a network incident, if the result of determining step (230) is a determination that sufficient information has been obtained to re-construct the road transport network incident from the image data.
4. A method of re-constructing a road transport network incident (200) according to Claim 1 or Claim 2, wherein the step of basing an operational road transport decision comprises carrying out further evidence - 22 gathering at a site of the network incident, if the result of determining step (230) is a determination that insufficient information has been obtained to re- construct the road transport network incident from the image data.
5. A method of re-constructing a road transport network incident (200) according to any preceding Claim, further characterised in that the step of determining (230) is performed automatically, based on processing the image data.
6. A method of re-constructing a road transport network incident (200) according to any preceding Claim, further characterised by the step of: :.

transferring image data (210, 220) of vehicles on: the road transport network from the array of camera units to a remote signal processing unit (160) that processes the image data to ascertain that a network incident has occurred. *S..

S S... S...
7. A method of re-constructing a road transport network incident (200) according to any preceding Claim further characterised by the step of: storing image data (215) relating to the vehicles obtained from the array of camera units so that the network incident can be subsequently reconstructed.
8. A method of re-constructing a road transport network incident (200) according to any preceding Claim further characterised in that the step of processing (225) the image data identifies a vehicle motion of one - 23 or more vehicles operating on the transport network at and/or prior to a time of the road transport network incident.
9. A method of re-constructing a road transport network incident (200) according to Claim 8, further characterised in that the identification of a vehicle motion enables calculation of a movement of a vehicle involved in the road transport network incident.
10. A method of re-constructing a road transport network incident (200) according to Claim 8 or Claim 9, further characterised in that the identification of a vehicle motion enables calculation of a speed of a vehicle involved in the road transport network incident.
11. A method of re-constructing a road transport network incident (200) according to any preceding Claim, further characterised in that the step of obtaining image SS* data (205) comprises obtaining image data from an array of camera units located on or adjacent to a highway routeing road transport.
12. A method of re-constructing a road transport network incident (200) according to any preceding Claim, further characterised by the step of configuring the array of camera units to capture overlapping images on a highway routeing road transport.
13. A method of re-constructing a road transport network incident (200) according to Claim 11 or Claim 12, further characterised in that the array of camera units - 24 - comprises two or more camera units located at points along a direction of traffic flow of the highway.
14. A method of re-constructing a road transport network incident (200) according to Claim 11 or Claim 12, further characterised by the step of configuring the array of camera units in a direction substantially perpendicular to a direction of traffic flow of the highway.
15. A method of re-constructing a road transport network incident (200) according to any of preceding Claims 11 to 14, further characterised by the step of configuring the array of camera units in substantially opposite directions.
16. A road transport network incident re-construction system (100) comprising: an array of camera units (150-158) for obtaining...

image data relating to a portion of a road transport network (110); and a signal processing unit (160), operably coupled...

to the array of camera units (150-158), for processing obtained image data; the road transport network incident re-construction system (100) characterised in that the signal processing unit (160) is adapted to determine whether sufficient information has been obtained to reconstruct a network incident based on the processed image data.
17. A road transport network incident re-construction system (100) according to Claim 16, further characterised - 25 - in that information outputs from the signal processing unit (160) determine an operational road transport decision of either clearing away a network incident or carrying out evidence-gathering at a site of the network incident.
18. A road transport network incident re-construction system (100) according to Claim 18, further characterised in that the signal processing unit (160) automatically determines an operational road transport decision based on the processed image data.
19. A road transport network incident re-construction system (100) according to any of preceding Claims 16 to 18, further characterised by a storage element, operably:* coupled to the signal processing unit, for storing image data relating to the road transport network incident obtained from the array of camera units. * * *
20. A road transport network incident re-construction system (100) according to any of preceding Claims 16 to 19, further characterised by the signal processing unit (160) processing the image data to identify a vehicle motion of one or more vehicles operating on the highway at and/or prior to a time of the road transport network incident.
21. A road transport network incident re-construction system (100) according to Claim 20, further characterised in that the identification of a vehicle motion enables calculation of a movement of a vehicle involved in the road transport network incident.

- 26 -
22. A road transport network incident re-construction system (100) according to Claim 20 or Claim 21, further characterised in that the identification of a vehicle motion enables calculation of a speed of a vehicle involved in the road transport network incident.
23. A road transport network incident re-construction system (100) according to any of preceding Claims 16 to 22 further characterised in that the array of camera units (150-158) is configured to capture overlapping images from respective camera units.
24. A road transport network incident re-construction system (100) according to any of preceding Claims 16 to:*. , 23 further characterised in that the array of camera units comprises two or more camera units (150-158) located at points along a direction of traffic flow of the highway.
25. A road transport network incident re-construction **..

system (100) according to any of preceding Claims 16 to 24 further characterised in that the array of camera units comprises two or more camera units located at points substantially perpendicular to a direction of traffic flow of the highway.
26. A road transport network incident re-construction system (100) according to any of preceding Claims 16 to 25 further characterised in that a number of camera units (150-158) are directed in substantially opposite directions.

-
27 - 27. A road transport network incident re-construction system (100) according to any of preceding Claims 16 to 26 further characterised in that the determination is a decision whether or not image data is of sufficient quality, and that this is performed on a basis of whether one or more image threshold(s) are exceeded.
28. A road transport network incident re-construction system (100) according to Claim 27 further characterised in that the one or more image threshold(s) comprise(s) an amount of data in each image that relates to a vehicle involved in the road transport network incident.
29. A road transport network incident re-construction;** * system (100) according to Claim 28 wherein each image is * : considered to be of sufficient quality if it contains image pixels that show more than a predetermined proportion of one or more of the vehicles involved in the i.'..

network incident. **S. **S.
30. A road transport network incident re-construction system (100) according to Claim 27 further characterised in that the determination of whether image data is of a sufficient quality is performed automatically by the processing unit analysing the captured image on a bit- wise basis.
31. A signal processing unit (160) adapted to perform the determining step (230) according to any of preceding Claims 1 to 15.

- 28 -
32. An array of cameras (150-158) for operably coupling to a processing unit (160) to determine a motion of one or more vehicles prior to, and/or during a traffic incident in accordance with the road transport network incident detection system according to any of preceding Claims 16 to 30.
33. A road transport network incident detection and/or re-construction system (100) substantially as hereinbefore described with reference to, and/or as illustrated by, FIG. 1 of the accompanying drawings.
34. A method of detecting and/or re-constructing a road transport network (200) incident substantially as hereinbefore described with reference to, and/or as:.

illustrated by, FIG. 2 of the accompanying drawings. * *

**. S. S * S *5*SSS S... * S S... S. *S a...