Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
A DRIVING AID SYSTEM AND METHOD OF CREATING A MODEL OF SURROUNDINGS OF A VEHICLE
Document Type and Number:
WIPO Patent Application WO/2007/114753
Kind Code:
A1
Abstract:
A driving aid system mounted in a road vehicle (1) and comprising: a detection system for detecting and storing profiles of characteristics of objects (2-9) around the vehicle along a road (10) on which the vehicle (1) is being driven; a positioning system for providing a current position of the vehicle (1), profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1); and a processing arrangement for comparing currently detected profiles with earlier stored profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle (1) relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1 ).

Inventors:
BAERGMAN JONAS (SE)
KAELLHAMMER JAN-ERIK (SE)
Application Number:
PCT/SE2007/000224
Publication Date:
October 11, 2007
Filing Date:
March 09, 2007
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
AUTOLIV DEV (SE)
BAERGMAN JONAS (SE)
KAELLHAMMER JAN-ERIK (SE)
International Classes:
G01C21/26; G05D1/02; G08G1/16
Domestic Patent References:
WO2005120882A12005-12-22
Foreign References:
US20050002558A12005-01-06
JP2004233220A2004-08-19
JPH11271074A1999-10-05
EP1536393A12005-06-01
Other References:
See also references of EP 2002210A4
Attorney, Agent or Firm:
MEULLER, Erik (Vårgårda, SE)
Download PDF:
Claims:

Claims

1. A driving aid system mounted in a road vehicle (1 ) and comprising: a detection system for detecting and storing profiles of characteristics of objects (2-9) around the vehicle (1) along a road (10) on which the vehicle (1 ) is being driven; a positioning system for providing a current position of the vehicle (1), profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1 ); and a processing arrangement for comparing currently detected profiles with earlier stored profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle (1) relative to an object (2-9) corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1 ), characterised by the positioning system comprising dead reckoning.

2. A driving aid system mounted in a road vehicle (1) and comprising: a detection system for detecting and storing profiles of characteristics of objects (2-9) around the vehicle (1) along a road (10) on which the vehicle (1) is being driven; a positioning system for providing a current position of the vehicle (1), profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1); and a processing arrangement for comparing currently detected profiles with earlier stored profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle (1 ) relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1 ),

characterised by the processing arrangement analysing the speed and direction of movement of moving objects (2-9) detected by the detection system.

3. A driving aid system according to claim 2, operable to determine whether a detected moving object (2-9) is a vehicle or a pedestrian, and storing the position and/or the movement path of the moving object.

4. A driving aid system according to claim 3, operable to inform a driver of the vehicle (1 ) if a pedestrian is detected in a location that is determined to be unusual.

5. A driving aid system according to claim 3 or 4, operable to inform a driver of the vehicle (1 ) if the detected behaviour of a pedestrian is determined to be unusual, with regard to the location of the pedestrian.

6. A driving aid system according to any one of claims 2 to 5, wherein the frequency or intensity of detected moving objects (2-9) along a movement path is stored.

7. A driving aid system according to claim 6, operable to identify areas of high pedestrian activity.

8. A driving aid system mounted in a road vehicle (1) and comprising: a detection system for detecting and storing profiles of characteristics of objects (2-9) around the vehicle (1) along a road (10) on which the vehicle is being driven; a positioning system for providing a current position of the vehicle (1), profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1 ); and

a processing arrangement for comparing currently detected profiles with earlier stored profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle (1) relative to an object (2-9) corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1), characterised by the detection system only storing profiles relating to objects (2-9) that appear to be fixed in position.

9. A driving aid system mounted in a road vehicle (1) and comprising: a detection system for detecting and storing profiles of characteristics of objects (2-9) around the vehicle (1) along a road (10) on which the vehicle (1) is being driven; positioning system for providing a current position of the vehicle (1), the positioning system comprising dead reckoning profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1); and a processing arrangement for comparing currently detected profiles with earlier stored profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle (1) relative to an object (2-9) corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1), characterised by profiles relating to defects along driving paths along which the vehicle (1) has not travelled for a predetermined time or distance being deleted.

10. A driving aid system according to any one of claims 2 to 9, wherein the positioning system comprises GPS or dead reckoning.

11. A driving aid system according to any preceding claim, wherein the positioning system comprises GPS and dead reckoning.

12. A driving aid system according to claim 11, wherein, if no GPS signal is received by the positioning system, the positioning system will use dead reckoning until a GPS signal is next received.

13. A driving aid system according to any preceding claim, wherein the detection system comprises a sensor arrangement for remotely sensing locations of objects (2-9) relative to the vehicle (1 ).

14. A driving aid system according to claim 13, wherein the detection system further comprises processing means for extracting simplified features of the objects (2-9) to provide said profiles, the simplified features of the objects (2-9) being added to a model of the surroundings of the vehicle (1 ).

15. A driving aid system according to any preceding claim, wherein allowable driving paths are determined from detected movement of the vehicle (1 ) and/or other vehicles.

16. A driving aid system according to any preceding claims, wherein, when a match between a currently detected profile and an earlier stored profile is found, the stored profile of the object (2-9) is enhanced and/or updated using the currently detected profile.

17. A driving aid system according to any preceding claim, wherein the driving aid system is operable to identify complex traffic areas.

18. A driving aid system according to claim 17, wherein profiles are only stored if they fall within areas classified as complex traffic areas.

19. A driving aid system according to Claim 17 or 18, wherein the driving aid system is operable to identify certain complex traffic areas as intersections.

20. A driving aid system according to claim 19, wherein intersections are classified as falling into one or more intersection classifications.

21. A driving aid system according to claim 20, wherein intersections are classified by comparing profiles of fixed and/or moving objects (2-9) against one or more pre-stored intersection templates.

22. A driving aid system according to claim 21 , operable to classify at least one type of intersection as the vehicle approaches the intersection, in the absence of any previously stored profiles or data relating to the intersection.

23. A driving aid system according to any one of claims 20 to 22, wherein the classification of an intersection is used to provide an indication to a driver of the vehicle (1) as to where the driver should look while approaching and/or negotiating the intersection.

24. A driving aid system according to any one of claims 17 to 21 wherein profiles are only stored if they fall within an areas that is determined to contain an intersection.

25. A driving aid system according to any one of claims 7 or 17 to 24, wherein at least some areas of high pedestrian activity or complex traffic areas are associated with times during which the areas are most busy.

26. A driving aid system according to any preceding claim, wherein profiles relating to objects (2-9) that are not present every time the vehicle passes the location in which the object (2-9) was previously detected are deleted.

27. A driving aid system according to any preceding claim, wherein the process of comparing earlier stored profiles with currently-detected profiles is initially conducted by comparing profiles relating to objects (2-9) closest to the vehicle (1).

28. A driving aid system according to any preceding claim, wherein the detection system is operable to make an initial determination of the current position of the vehicle (1) when the vehicle (1) approaches an area containing one or more objects (2-9) in respect of which profiles have been stored, and wherein the position of the vehicle (1) relative to one or more objects (2-9) in respect of which profiles are stored is determined when a match between a currently detected profile and a stored profile is made.

29. A driving aid system according to any preceding claim, which is operable to make a determination that a hazardous situation may occur or is likely to occur, and to activate one or more vehicle safety systems in response to the making of such a determination.

30. A driving aid system according to claim 29, wherein the manner in which one or more vehicle safety systems is activated is varied in dependence upon information provided by the driving aid system.

31. A driving aid system according to claim 29 or 30, wherein activation of the vehicle safety systems causes a warning to be presented to the driver.

32. A driving aid system according to any one of claims 28 to 31 , wherein activation of the vehicle safety system causes emergency braking of the vehicle (1) to be initiated, and/or a seat belt pretensioner to be activated.

33. A driving aid system according to any preceding claim, wherein the position of the vehicle (1 ) relative to a stored map is determined, with the position of the vehicle (1) relative to said map being corrected and/or updated when a match between a stored profile and a currently-detected profile is made.

34. A method of creating a model of surroundings of a vehicle (1), comprising: detecting and storing profiles of characteristics and objects (2-9) around the vehicle (1 ) along a road (10) on which the vehicle (1 ) is being driven; providing a current position of the vehicle (1) using a positioning system, profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1); comparing currently detected profiles with earlier stored profiles; and if a match between a currently detected profile and an earlier profile is found, determining the current position of the vehicle (1), relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1 ), characterised by the positioning system comprising dead reckoning.

35. A method of creating a model of surroundings of a vehicle (1 ), comprising: detecting and storing profiles of characteristics and objects (2-9) around the vehicle (1) along a road (10) on which the vehicle (1) is being driven; providing a current position of the vehicle (1) using a positioning system, profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1); comparing currently detected profiles with earlier stored profiles; and if a match between a currently detected profile and an earlier profile is found, determining the current position of the vehicle (1), relative to an object

corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1 ), characterised by the step of analysing the speed and direction of detected moving objects.

36. A method of creating a model of surroundings of a vehicle (1), comprising: detecting and storing profiles of characteristics and objects (2-9) around the vehicle (1) along a road (10) on which the vehicle (1) is being driven; providing a current position of the vehicle (1) using a positioning system, profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1); comparing currently detected profiles with earlier stored profiles; and if a match between a currently detected profile and an earlier profile is found, determining the current position of the vehicle (1), relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1 ), characterised by the step of storing profiles relating only to objects that appear to be fixed in position.

37. A method of creating a model of surroundings of a vehicle (1), comprising: detecting and storing profiles of characteristics and objects (2-9) around the vehicle (1) along a road (10) on which the vehicle (1) is being driven; providing a current position of the vehicle (1) using a positioning system, profiles of characteristics of objects (2-9) being stored in relation to the detected position of the vehicle (1); comparing currently detected profiles with earlier stored profiles; and if a match between a currently detected profile and an earlier profile is found, determining the current position of the vehicle (1), relative to an object

corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle (1 ), characterised by the step of deleting profiles relating to objects along driving paths along which the vehicle has not travelled for a predetermined time or distance

38. A method according to any one of claims 34 to 37, further comprising the steps of: determining that a hazardous situation may occur or is likely to occur; and activating one or more vehicle safety systems in response to the making of such a determination.

39. A computer program comprising computer program code operable to perform all of the steps of any one of claims 34 to 38 when said program is run on a computer.

40. A computer program according to claim 39, embodied on a computer readable medium.

Description:

A driving aid system and method of creating a model of surroundings of a vehicle

Description of Invention

THIS INVENTION relates to a driving aid system, and in particular concerns a driving aid system to be mounted in a road vehicle for increased vehicle safety and accuracy in determining the position of the vehicle.

For road vehicles such as cars, it is known to provide a navigation system (such as GPS) which, along with stored map data, may give information regarding speed limits, intersections, traffic lights and so forth, so that the driver can be warned when it is determined that the vehicle may encounter a hazardous situation. It is also known to use "dead reckoning" to determine the current position of the vehicle relative to an earlier known position, if, for example, a GPS signal is lost (such a system may use accelerometers, gyroscopes and/or wheel rotation sensors to track the movements of the vehicle). It will be appreciated, however, that the further the vehicle progresses from the last known position, the degree of accuracy with which the position of the vehicle can be determined will steadily reduce.

It is an object of the present invention to seek to provide an improved system for determining the position of the vehicle, and also for alerting the driver of the vehicle to potential hazards.

Accordingly, one aspect of the present invention provides a driving aid system mounted in a road vehicle and comprising: a detection system for detecting and storing profiles of characteristics of objects around the vehicle along a road on which the vehicle is being driven; a positioning system for providing a current position of the vehicle, profiles of characteristics of objects being stored in relation to the detected position of the vehicle; and a processing arrangement for comparing currently detected profiles with earlier stored

profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle, characterised by the positioning system comprising dead reckoning.

Another aspect of the present invention provides a driving aid system mounted in a road vehicle and comprising: a detection system for detecting and storing profiles of characteristics of objects around the vehicle along a road on which the vehicle is being driven; a positioning system for providing a current position of the vehicle, profiles of characteristics of objects being stored in relation to the detected position of the vehicle; and a processing arrangement for comparing currently detected profiles with earlier stored profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle, characterised by the processing arrangement analysing the speed and direction of movement of moving objects detected by the detection system.

Advantageously, the driving aid system is operable to determine whether a detected moving object is a vehicle or a pedestrian, and storing the position and/or the movement path of the moving object.

Preferably, the driving aid system is operable to inform a driver of the vehicle if a pedestrian is detected in a location that is determined to be unusual.

Conveniently, the driving aid system is operable to inform a driver of the vehicle if the detected behaviour of a pedestrian is determined to be unusual, with regard to the location of the pedestrian.

Advantageously, the frequency or intensity of detected moving objects along a movement path is stored.

Preferably, the driving aid system is operable to identify areas of high pedestrian activity.

A further aspect of the present invention provides a driving aid system mounted in a road vehicle and comprising: a detection system for detecting and storing profiles of characteristics of objects around the vehicle along a road on which the vehicle is being driven; a positioning system for providing a current position of the vehicle, profiles of characteristics of objects being stored in relation to the detected position of the vehicle; and a processing arrangement for comparing currently detected profiles with earlier stored profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle, characterised by the detection system only storing profiles relating to objects that appear to be fixed in position.

Another aspect of the present invention provides a driving aid system mounted in a road vehicle and comprising: a detection system for detecting and storing profiles of characteristics of objects around the vehicle along a road on which the vehicle is being driven; positioning system for providing a current position of the vehicle, the positioning system comprising dead reckoning profiles of characteristics of objects being stored in relation to the detected position of the vehicle; and a processing arrangement for comparing currently detected profiles with earlier stored profiles and, if a match between a currently detected profile and an earlier stored profile is found, determining the current position of the vehicle relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future

surroundings of the vehicle, characterised by profiles relating to defects along driving paths along which the vehicle has not travelled for a predetermined time or distance being deleted.

Conveniently, the positioning system comprises GPS or dead reckoning.

Advantageously, the positioning system comprises GPS and dead reckoning.

Preferably, if no GPS signal is received by the positioning system, the positioning system will use dead reckoning until a GPS signal is next received.

Conveniently, the detection system comprises a sensor arrangement for remotely sensing locations of objects relative to the vehicle.

Advantageously, the detection system further comprises processing means for extracting simplified features of the objects to provide said profiles, the simplified features of the objects being added to a model of the surroundings of the vehicle.

Preferably, allowable driving paths are determined from detected movement of the vehicle and/or other vehicles.

Conveniently, when a match between a currently detected profile and an earlier stored profile is found, the stored profile of the object is enhanced and/or updated using the currently detected profile.

Advantageously, the driving aid system is operable to identify complex traffic areas.

Preferably, profiles are only stored if they fall within areas classified as complex traffic areas.

Conveniently, the driving aid system is operable to identify certain complex traffic areas as intersections.

Advantageously, intersections are classified as falling into one or more intersection classifications.

Preferably, intersections are classified by comparing profiles of fixed and/or moving objects against one or more pre-stored intersection templates.

Conveniently, the driving aid system is operable to classify at least one type of intersection as the vehicle approaches the intersection, in the absence of any previously stored profiles or data relating to the intersection.

Advantageously, the classification of an intersection is used to provide an indication to a driver of the vehicle as to where the driver should look while approaching and/or negotiating the intersection.

Preferably, profiles are only stored if they fall within an areas that is determined to contain an intersection.

Conveniently, at least some areas of high pedestrian activity or complex traffic areas are associated with times during which the areas are most busy.

Advantageously, profiles relating to objects that are not present every time the vehicle passes the location in which the object was previously detected are deleted.

Preferably, the process of comparing earlier stored profiles with currently- detected profiles is initially conducted by comparing profiles relating to objects closest to the vehicle.

Conveniently, the detection system is operable to make an initial determination of the current position of the vehicle when the vehicle approaches an area containing one or more objects in respect of which profiles have been stored, and wherein the position of the vehicle relative to one or more objects in respect of which profiles are stored is determined when a match between a currently detected profile and a stored profile is made.

Advantageously, the driving aid system is operable to make a determination that a hazardous situation may occur or is likely to occur, and to activate one or more vehicle safety systems in response to the making of such a determination.

Preferably, the manner in which one or more vehicle safety systems is activated is varied in dependence upon information provided by the driving aid system.

Conveniently, activation of the vehicle safety systems causes a warning to be presented to the driver.

Advantageously, activation of the vehicle safety system causes emergency braking of the vehicle to be initiated, and/or a seat belt pretensioner to be activated.

Preferably, the position of the vehicle relative to a stored map is determined, with the position of the vehicle relative to said map being corrected and/or updated when a match between a stored profile and a currently-detected profile is made.

A further aspect of the present invention provides a method of creating a model of surroundings of a vehicle, comprising: detecting and storing profiles

of characteristics and objects around the vehicle along a road on which the vehicle is being driven; providing a current position of the vehicle using a positioning system, profiles of characteristics of objects being stored in relation to the detected position of the vehicle; comparing currently detected profiles with earlier stored profiles; and if a match between a currently detected profile and an earlier profile is found, determining the current position of the vehicle, relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle, characterised by the positioning system comprising dead reckoning.

Another aspect of the present invention provides a method of creating a model of surroundings of a vehicle, comprising: detecting and storing profiles of characteristics and objects around the vehicle along a road on which the vehicle is being driven; providing a current position of the vehicle using a positioning system, profiles of characteristics of objects being stored in relation to the detected position of the vehicle; comparing currently detected profiles with earlier stored profiles; and if a match between a currently detected profile and an earlier profile is found, determining the current position of the vehicle, relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle, characterised by the step of analysing the speed and direction of detected moving objects.

A further aspect of the present invention provides a method of creating a model of surroundings of a vehicle, comprising: detecting and storing profiles of characteristics and objects around the vehicle along a road on which the vehicle is being driven; providing a current position of the vehicle using a positioning system, profiles of characteristics of objects being stored in relation to the detected position of the vehicle; comparing currently detected profiles with earlier stored profiles, and if a match between a currently detected profile and an earlier profile is found, determining the current position of the vehicle,

relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle, characterised by the step of storing profiles relating only to objects that appear to be fixed in position.

Another aspect of the present invention provides a method of creating a model of surroundings of a vehicle, comprising: detecting and storing profiles of characteristics and objects around the vehicle along a road on which the vehicle is being driven; providing a current position of the vehicle using a positioning system, profiles of characteristics of objects being stored in relation to the detected position of the vehicle; comparing currently detected profiles with earlier stored profiles, and if a match between a currently detected profile and an earlier profile is found, determining the current position of the vehicle, relative to an object corresponding to the earlier stored profile, and using the same or other earlier stored profiles to predict the future surroundings of the vehicle, characterised by the step of deleting profiles relating to objects along driving paths along which the vehicle has not travelled for a predetermined time or distance.

Conveniently, the method further comprises the steps of: determining that a hazardous situation may occur or is likely to occur; and activating one or more vehicle safety systems in response to the making of such a determination.

A further aspect of the present invention provides a computer program comprising computer program code operable to perform all of the steps of the above when said program is run on a computer.

Another aspect of the present invention provides a computer program according to the above, embodied on a computer readable medium.

In order that the present invention may be more readily understood, embodiments thereof will now be described, by way of example, with reference to the accompanying drawings, in which:

FIGURE 1 is a flow chart showing steps carried out by a driving aid system embodying the present invention;

FIGURES 2 and 3 show primary and secondary profiles respectively, created by a driving aid system embodying the present invention, of an area through which a vehicle containing the driving aid system is progressing; and

FIGURE 4 shows a model produced by a driving aid system embodying the present invention.

A driving aid system embodying the present invention is mounted inside a road vehicle, for instance a car. The driving aid system comprises a detection system, which is operable to detect and store profiles of characteristics of objects around the vehicle, along a road on which the vehicle is being driven. The detection system is preferably able to detect profiles of characteristics of objects around the vehicle remotely and may, for example, comprise a horizontally scanning laser radar, which measures the time taken for an emitted laser signal to return and, combined with the direction of return of the emitted signal, determines information regarding characteristics (such as position, shape, size and direction and speed of travel) of an object from which the signal has been reflected. Any suitable detection system may be used, and it is envisaged that the present invention may also function with conventional radar, optical or ultrasonic systems.

A memory (which may be a part of the detection system) is provided to allow detected profiles of characteristics of objects around the vehicle to be stored and subsequently accessed.

The driving aid system further comprises a positioning system for providing a current position of the vehicle. An example of such positioning system is the Global Positioning System (GPS), although any other suitable positioning system may be used.

Use of a driving aid system embodiment of the present invention will now be described.

As a vehicle in which the driving aid system is mounted is driven along a road, the detection system detects profiles of characteristics of objects around the vehicle, as described above. Processes for determining the characteristics of the surrounding objects from the data received by the detection system are known, and will not be discussed in this specification.

The profiles of characteristics of objects around the vehicle are stored, and when they are stored the profiles are associated or tagged with an absolute position, which is determined by calculating the position of the object relative to the vehicle, and from the position of the vehicle as determined by the positioning system. In this way, it will be appreciated that, as a vehicle containing the driving aid system is driven between various locations, profiles relating to objects around the driving path of the vehicle will be detected and stored. Profiles may be buffered for a certain length of time before being stored, to allow a decision to be taken as to whether they should indeed be stored. For instance (as discussed in more detail below) storage may only be required if the object in question forms part of an intersection or other area of interest and on first detection of an object it may not be clear whether this is the case. Buffering profiles allow time for an appropriate determination to be made, so that the profile can be correctly stored or discarded.

If the vehicle is driven along a driving path along which it has already been driven on one or more previous occasions, then profiles of characteristics of objects along the driving path may have already been detected and stored in the memory. As the vehicle progresses, the driving aid system retrieves the stored profiles of objects that were detected during the one or more previous journeys, and attempts to match profiles detected by the detection system to the profiles that have already been stored, by comparing the detected position and/or other characteristics of surrounding objects with the stored profiles. In doing so, the driving aid system uses the currently determined position of the vehicle as a starting point, and from the absolute position data that is associated with the previously-stored profiles, makes a determination as to where the previously-detected objects should be relative to the position of the vehicle. Initially, the driving aid system may attempt to find matches relating to objects closest to the vehicle for which profiles have been previously stored.

If the detection system detects characteristics of a nearby object that matches stored characteristics from one or more previous journeys, the driving aid system may make a match between the stored profile and the currently detected profile, and proceed to update and/or enhance the profile of the object in the database using the newly-acquired data from the detection system. It will be appreciated that detecting the characteristics of an object in this manner on more than one occasion will improve the accuracy with which the object's characteristics (e.g. size, position and shape) may be determined.

It will also be appreciated that, if an object is detected during one journey but not detected during another, it may be determined that the object is not a permanent feature, and the object may therefore be removed from the memory. For instance, a parked car and a lamp post may be detected during a first journey on a driving path, and on a subsequent journey the car may not be detected (because it has been moved), whereas the lamp post will be detected as before. In this case, the profile corresponding to the parked car

will be removed from the memory, whereas the profile of the lamp post will be retained, and may be updated and/or enhanced using the newly-acquired data.

Amongst the fixed objects, the driving aid system may determine the positions of possible driving paths (e.g. roads), and also retain this information. The position of driving paths may be determined, for example, by keeping a record of the routes along which the vehicle is driven, and also by detecting and noting the movement of other vehicles. In such embodiments of the invention, the detection system is provided with an algorithm that is able to distinguish another road vehicle from other types of objects, and any route along which a road vehicle is seen to travel, or along which a predetermined number of vehicles are seen to travel, may be designated as a possible driving path.

If a sufficient number of passes are made along a particular road, and other road vehicles are encountered, the driving aid system may be able to determine the number of lanes that are provided on a particular road, and also the direction of permitted travel of road vehicles along each lane.

The driving aid system may also be operable to identify regions having consistently heavy traffic activity, or where the detected directions of travel of the vehicle and of other road vehicles vary in such a way that it may be determined that that region comprises complex traffic area, which may be a road intersection. For instance, if, on a particular driving path, the detected directions of travel of the vehicle and of other road vehicles comprise only travel in one direction and travel in a substantially opposite direction, it might reasonably be determined that this driving path comprises a straight stretch of road, with no intersection or side branching. However, if it is determined that the vehicle sometimes passes along a particular driving path in a first direction, but on other occasions passes across the driving path in a direction which is substantially at right angles to the first direction, or if other road

vehicles are observed travelling across the driving path in a direction which is substantially at right angles to the first direction, then it may be determined that the region in question comprises a junction or intersection.

In embodiments of the present invention, the driving aid system is able to identify and store the positions of junctions and/or intersections, or other regions in which the movement of the traffic is determined to be heavy or complex.

The driving aid system may also be operable to classify the type of junction or intersection that has been detected, for instance a cross-junction, a T-junction, a flyover and so on. It is envisaged that this may be achieved by storing "template" characteristics of different types of junctions and intersections, and classify a junction or intersection according to template characteristics that are most accurately matched by the observed traffic behaviour. The junctions or intersections may also be classified with respect to the positions of fixed objects in conjunction with the traffic behaviour, for instance the number of lanes in different directions and physical features such as traffic lights, as well as associated driving behaviour.

On approaching a new junction or intersection for the first time, the driving aid system may be operable to classify the junction or intersection before the arrival of the vehicle at the junction or intersection. In this case, the driving aid system may be in a position to take an appropriate action, for instance the activation of the vehicle safety system (see below for a more detailed discussion), as appropriate for the type of junction or intersection and/or detected moving or stationary objects.

The driving aid system may also be provided with one or more algorithms allowing the driving system to determine whether a detected object is likely to be a pedestrian. From the detected behaviour (e.g. positions and directions of

travel) of pedestrians, the driving aid system may be able to determine the positions of pedestrian routes around driving paths, for instance pavements, paths, pedestrian crossings, traffic lights, and so on. Features where many pedestrians are detected may once again be classified by the driving aid system, and again it is envisaged that this may be done by comparing pedestrian behaviour to one or more stored templates, each of which corresponds to a particular type of pedestrian feature. The driving aid system may also identify pedestrian features as being busy, if a high amount of pedestrian activity is detected on one occasion or is consistently detected. On approaching such areas, the driver may be given a warning, and threshold levels for certain safety systems (e.g. emergency braking) may be altered, particularly to make the safety systems more sensitive.

Regions designated as junctions, intersections, areas of high pedestrian activity and so on may also be associated with times at which this activity is greatest. For instance, the driving aid system may identify that there is high pedestrian activity outside a school at certain hours of the day, and on weekdays (as opposed to weekends). Warnings may therefore be given to the driver of the vehicle as it approaches the school at the appropriate time, or on an appropriate day, but not at other times. Similarly, the driving aid system may determine that an intersection is busy at certain times (e.g. during morning and evening rush hours) but is seen as significantly less busy at other times.

Also, if a pedestrian is detected in an unusual place, or where pedestrians have never or seldom been detected before, or if the detected behaviour of a pedestrian is determined to be unusual (particularly considering the location of the pedestrian), the driver may be warned of this. In embodiments of the invention, information regarding pedestrians (and indeed regarding other moving or stationary objects) may be received from other vehicle subsystems, such as one or more video or infrared cameras.

It is also envisaged that previously-stored profiles may be used to determine the current position of the vehicle with greater accuracy. If, for example, a particular object has been detected several times, the driving aid system may store the position of the object with a high degree of accuracy (it is envisaged that the accuracy may be to within less than 0.1 metres). In contrast, at any given moment the position of the vehicle may be known only to a relatively low degree of accuracy from the positioning system (typically greater than 0.1 metres). In addition, if a signal required to determine the position of the vehicle has been lost, then the vehicle may be attempting to keep a track of the current position using dead reckoning, and, as discussed above, the accuracy with which the current position can be determined by this method rapidly decreases with time and distance travelled. It will be appreciated that such a stored map contains different information to that contained in the model built up by the driving aid system, but it is envisaged that there will be sufficient features in common for such matches to be made.

If, however, the vehicle is able to detect the position, relative to itself, of an object whose position is stored with a high degree of accuracy in the memory, then the vehicle will be able to determine its current position accurately. If a driving aid system is able to make an accurate determination of this nature, then the current position of the vehicle, as determined by the positioning system may be updated, and a correction made for any offset that may be present between the current position as determined by the positioning system and the actual position of the vehicle. The position of the vehicle on a stored map may also be updated or corrected using this method, in particular by relating the stored profiles to features present on the map. It will be appreciated that such a stored map will contain different information to that contained in the model built up by the driving aid system, but it is envisaged that there will be sufficient features in common for such matches to be made.

It will be appreciated that, if the vehicle is driven over large distances, the quantity of data that must be stored to maintain an accurate set of profiles for objects surrounding the driving paths will be very large. In embodiments of the present invention, the creation of maps of detected objects, intersections and so on is concentrated on driving paths along which the vehicle is driven frequently. This has the benefit of concentrating the memory resources on such areas, and therefore, on locations in which the vehicle is most likely to encounter a hazardous situation.

In such embodiments of the invention, stored profiles relating to driving paths along or through which the vehicle has not been driven for a predetermined time (for instance one month or one year) or distance (for instance 1 ,000, 10,000 or 20,000 km) may be deleted from the memory. Profiles along a driving path may be retained for a longer time if the vehicle has passed along the driving path more than a predetermined number of times. The driving aid system may additionally retain in memory for a longer time profiles relating to locations which may present hazardous situations, for instance intersections or pedestrian features. Indeed, in some embodiments the driving aid system may retain only profiles relating to intersections and/or locations determined to be "complex" traffic areas, or otherwise areas of potential hazard. If the available storage memory becomes full, then the "oldest" profiles may be deleted as and when storage space is required for the storage of new data, for instance relating to a new complex traffic area that has been identified.

In use of the driving aid system, the driving aid system uses the current position of the vehicle as determined by the positioning system and/or by matches with previously-stored profiles, and (taking into account the direction of travel of the vehicle) predicts the future surroundings of the vehicle, by extrapolating the motion of the vehicle and thus determining which objects, intersections, etc. will be encountered.

The driving aid system preferably uses one or more algorithms to determine whether the vehicle is likely to encounter a hazardous situation, and to activate one or more appropriate safety system of the vehicle if such a determination is made.

For instance, if the driving aid system determines that the vehicle is being driven towards an intersection which is often busy, a warning may be presented to the driver of the vehicle, and this warning may include information regarding the classification of the intersection, and/or information regarding the usual rate of traffic at the intersection, and/or indicate to the driver where the driver should look when approaching and negotiating the intersection to notice potential hazards. In some embodiments, the warning may only be presented or a relatively strong warning may be presented when it is determined that the speed and/or direction of travel of the vehicle on approaching the intersection are such that action will need to be taken swiftly if the intersection is to approached and/or negotiated safely.

If it appears necessary, information regarding the position of the vehicle may be transmitted to other vehicles, or via a communications network to an external location.

In extreme examples, if the speed and/or direction of travel of the vehicle as the intersection is approached are such that a hazardous situation appears to be likely, further vehicle safety systems may be automatically activated such as emergency braking, brake assist, steering assist, steering avoidance, a reversible seat belt pretensioner and/or other safety devices.

The determination as to whether the vehicle has encountered, or is likely to encounter, a hazardous situation, is made not only from the position and speed of the vehicle, but may also be made in response to the detection of non-permanent moving or non-moving objections. For instance, although a

driver may be negotiating an intersection in an appropriate, but the presence of an unusual fixed object (for instance a large object which is lying on the road) or unusual moving objections (for instance another vehicle which is driving hazardously, or a pedestrian who is, for some reason, standing on the road) will lead the system to determine that a potential hazardous situation may arise, or indeed has arisen.

In further embodiments of the invention, the driving aid system predicts the future surroundings of the vehicle over only a short distance, for example around 100m - 400m. It is appreciated that this will reduce the burden on the processing capabilities of the driving aid system. In advantageous embodiments of the invention, the distance over which the future surroundings of the vehicle is predicted is varied in dependence upon the speed with which the vehicle is travelling.

With reference to Figure 2, a vehicle 1 is shown, travelling through an environment in respect of which no data is already stored by the driving aid system. Shown around the vehicle 1 are "primary profiles" of regions from which it appears that signals emitted by the detection system have been reflected and returned.

The primary profiles are then used to create "secondary" profiles of objects 2 to 9, (as shown in Figure 3) which it is determined are present around the vehicle.

Turning now to Figure 4, after several passes through the environment, the driving aid system may build up a stored model of the environment, which includes the locations of the roads 10 as allowable driving paths (along with the number of lanes at each point along the road, and the permitted direction of travel in each lane) as well as fixed objects which appear between the driving paths. When driving through this environment again, the distance

between the vehicle 1 and each of the known objects may be determined, as well as the position of the vehicle 1 relative to the roads 10, and the lanes 11 within the roads 10. The driving aid system is then able to determine whether the vehicle 1 is travelling in an appropriate direction with regard to the lane 11 in which the vehicle 1 is currently travelling, whether the vehicle 1 is on a collision course of any of the fixed objects 2-9 or whether the vehicle 1 is approaching a junction, intersection or pedestrian feature, as discussed above. The driving aid's model of the surroundings may, as indicated above, also use information from other vehicle subsystems, such as one or more video cameras and/or infrared cameras, particularly for detecting and analysing the behaviour of moving objects, e.g. pedestrians. If a determination is made that a hazardous situation may be or is likely to occur, appropriate safety measures (as discussed above) may be taken.

If safety systems such as emergency braking, emergency steering and deployment of an air-bag are activated, then the manner of the activation may depend upon the information gathered by the driving aid system. For instance, the severity of the emergency braking or the direction of the emergency steering may be determined by the locations of moving and/or stationary objects detected whilst negotiating an intersection. As a further example, thresholds for the activation of safety systems (e.g. detected distance to a stationary object) may be raised or lowered, or indeed certain safety systems may be activated or deactivated, depending on the information provided by the driving aid system.

It will be appreciated that embodiments of the present invention may improve the overall level of safety when driving a vehicle, and also allow more accurate determination of the position of the vehicle.

When used in this specification and claims, the terms "comprises" and "comprising" and variations thereof mean that the specified features, steps or

integers are included. The terms are not to be interpreted to exclude the presence of other features, steps or components.