Method of detecting an incident or the like on a portion of a route

• US 8,055,015 B2
• Filed: 07/11/2003
• Issued: 11/08/2011
• Est. Priority Date: 07/22/2002
• Status: Active Grant

• Current Assignee: Citilog Co.
• Original Assignee: Citilog Co.
• Inventors: Bouzar, Salah
• Primary Examiner(s): Desire; Gregory M
• Assistant Examiner(s): Allison; Andrae S
• Application Number:
  US10/521,606
• Publication Number:
  US 20050213791A1
• Time in Patent Office: 3,042 Days
• Field of Search: 382/103, 382/104, 382/154, 340/903, 340/928, 340/933, 340/988, 701/117, 701/201, 348/148, 348/149, 348/164, 348/169
• US Class Current: 382/103
• CPC Class Codes: G08G 1/04 using optical or ultrasonic detectors

Abstract

A method for detecting an incident on a traffic lane uses a video camera having a target constituting an optoelectronic converter of a real optical image, the target being monitored by a programmable processing member and the incident detection process being designed to be carried out by activating the programmable processing member only when the real landscape image focused on the target is stationary. The inventive method includes detecting the beginning of displacement of the real image of the landscape relative to the target, in deactivating the programmable processing member right at the beginning of the displacement of the real image of the landscape relative to the target, and reactivating the programmable processing member at the end of the displacement of the real image of the landscape relative to the target to carry out the incident detection process.

Claims

1. A method of detecting an incident on a portion of route situated in a scene said portion of route is suitable for having objects travelling therealong, and the method using a video camera controllable in one of azimuth, elevation and field of view and having a target constituting an optoelectronic converter of a real optical image of the scene, said target being controlled by a programmable processor member, said process detecting incidents comprising:

A) having said programmable processing member determine that said video camera is substantially stationary in relation to said scene by

i) selecting at least one point on a current real image of said scene, outside of said portion of said route;

ii) comparing said at least one point on said current real image with said at least one point on at least one of a set of immediate previous targets of a stored image;

iii) determining that said at least one point is approximately at the same position in said current real image as in at least one of the set of immediate previous targets of the stored image;

iv)a) upon failure to determine the same position in step iii, re-execute steps i) through iii) with at least one other point; and

iv)b) upon determining the same position in step iii), proceed to step B)

B) having said programmable processor member process said current real images to detect incidents.

2. The method according to claim 1, wherein the real image of the scene beginning to move relative to the target occurs upon the beginning of a zooming in function or a zooming out function of the real image.

3. The method according to claim 1, wherein the end of the movement of the real image of the scene relative to the target occurs upon an end of a zooming in function or a zooming out function of the real image.

4. The method of claim 1, wherein the programmable processor member is deactivated as soon as the stationarity of the scene relative to the target is detected as ending, and reactivated, in order to implement the process for detecting an incident, as soon as said stationarity of the scene relative to the target is detected as beginning.

5. The method of claim 4, wherein said beginning and ending of said stationarity of the scene relative to the target are detected by determining at least one first image point of said real image of the scene corresponding to a stationary point of said scene, substantially outside said portion of said route, by generating a first command signal when said first image point is subjected to a change of position of said target, and by controlling said programmable processor member as a function of said first command signal.

6. The method of claim 5, wherein beginning and ending of said stationarity of the scene relative to the target are detected by determining at least one second and one third image points of said real image of the scene corresponding respectively to two stationary points of said scene, substantially outside said portion of said route, by generating a second command signal when distance between said second and third image points changes, and by controlling said programmable processor member as a function of said second command signal.

7. The method of claim 6, wherein said beginning and ending of stationarity of the scene relative to the target are detected by determining at least one fourth and on fifth image points of said real image of the scene corresponding respectively to two stationary points of said scene, substantially outside said portion of said route, by generating a third command signal when distance between said fourth and fifth image points changes, and by controlling said programmable processor member as a function of said third command signal.

8. The method of claim 7, wherein said beginning and ending of stationarity of the scene relative to the target are detected by determining a plurality of image points of said real image of the scene corresponding to a plurality of points which are stationary at the beginning of movement of said real image of the scene and substantially on said portion of the route, by generating a fourth command signal when a determined number of said plurality of image points have become stationary again at the end of movement of said real image of the scene, and by controlling said programmable processor member as a function of said fourth command signal.

9. The method of claim 5, wherein said target is subdivided into a plurality of photosensitive points suitable for delivering signals as a function of the quantity of radiation received by their photosensitive surfaces.

10. The method of claim 6, wherein said target is subdivided into a plurality of photosensitive points suitable for delivering signals as a function of the quantity of radiation received by their photosensitive surfaces.

11. The method of claim 7, wherein said target is subdivided into a plurality of photosensitive points suitable for delivering signals as a function of the quantity of radiation received by their photosensitive surfaces.

12. The method of claim 1, wherein said target is subdivided into a plurality of photosensitive points suitable for delivering signals as a function of the quantity of radiation received by their photosensitive surfaces.

13. The method of claim 12, wherein the process for detecting an incident on said portion of route when it is suitable for having objects traveling thereon along an axis and following a path that is substantially imposed, comprises:

in selecting a group of photosensitive points in said plurality of photosensitive points of the target, the selected group of points corresponding to points of said portion of route located on a plurality of main geometrical construction lines, said main construction lines being situated in the plane of said portion of route and all being substantially parallel to the axis of said trajectory; and

in analyzing the set of signals delivered by the photosensitive points of said selected group.

14. The method of claim 13, wherein the detection process further comprises:

in subdividing said selected group of photosensitive points into a plurality of subgroups of photosensitive points corresponding to points on the portion of route situated at the intersections between said main construction lines and respective secondary geometrical construction lines that are substantially perpendicular to the main construction line; and

in associating each photosensitive point of a subgroup with a weighting coefficient for multiplying the value of the signal emitted by said point, the weighting coefficient being a function of the preferential probability of objects passing over the point of the portion of route whose image is the photosensitive point associated with said weighting coefficient.

15. The method of claim 14, wherein the photosensitive receive areas of said photosensitive points are of substantially the same dimensions.

16. The method of claim 15, wherein said analysis further comprises:

averaging the values of the signals delivered by the points of each subgroup at given instants;

comparing the averages as obtained in this way for each subgroup; and

deducing from said comparison the presence, if any, of an incident on said portion of route.

17. An automatic road traffic incident detection system comprising at least one video camera with an optical axis controllable in azimuth, elevation and focal distance, said camera being positioned alongside said road and fit for taking real images of scenes of the road and converting them into target images which are submitted as an entry to a computer process to detect traffic incidents, said computer process being deactivated by a programmable processing member based on detection of movement of the at least one video camera by the ending of stationarity of said target images relative to said real images of scenes and being reactivated based on detection of lack of movement of the at least one video camera by the beginning of stationarity of said target images relative to real images of scenes, wherein said ending and beginning of stationarity of said target images relative to said real images of scenes are detected by selecting at least one point on a current real image of said scenes, outside of said portion of said route, comparing said point on said current real image with said point on at least one of a set of immediate previous targets of a stored image, determining that stationarity begins upon said point being approximately at the same position in said current real image and said set of immediate previous targets of a stored image relative to said scene and that stationarity ends upon said point not being in the same position in said current real image and said set of immediate previous targets relative to said scene.

18. A system for detecting an incident in the flow of traffic, comprising:

a camera with an optical axis controllable in azimuth, elevation and focal distance for taking images of roadside scenes and transmitting them for determination of a roadside incident; and

a programmable processor programmed to detect the roadside incident from images received from the camera,

wherein the programmed processor detects when the roadside incident processing ends based upon a determination that the camera is moving relative to a scene, the determination of camera movement based upon changing points in a current live image relative to a set of previous images,

the programmed processor detects the roadside incident processing begins based upon a determination that the camera is stationary relative to the scene, the determination of camera movement based upon points remaining fixed in the current live image relative to the set of previous images.