Method for identifying objects in motion, in particular vehicles, and systems for its implementation

• US 5,083,200 A
• Filed: 04/02/1990
• Issued: 01/21/1992
• Est. Priority Date: 03/31/1989
• Status: Expired due to Fees

• Current Assignee: Elsydel (Ajax Holding SAS)
• Original Assignee: Elsydel (Ajax Holding SAS)
• Inventors: Deffontaines, Thierry
• Primary Examiner(s): Peng, John K.
• Application Number:
  US07/502,878
• Time in Patent Office: 659 Days
• Field of Search: 358/108, 358/105, 358/106, 358/93, 358/107, 358/101, 358/96, 340/937, 340/942, 356/23, 356/26, 356/237, 364/560, 364/562, 250/222.1
• US Class Current: 348/148
• CPC Class Codes: G07B 15/063 using wireless information transmission between the vehicle and a fixed station identification of vehicles G08G
• CPC Class Codes: G08G 1/04 using optical or ultrasonic detectors

Abstract

A method for identifying an object in motion, in particular a vehicle, includes several steps whenever the object is moving inside a predetermined identification zone following a predetermined movement axis. The steps are periodically acquiring images of the object in a predetermined field of view, checking the nature of the image background in the field of view to obtain background reference information in the absence of the object, and processing the images acquired in combination with the background reference information in order to extract therefrom a silhouette of the object having crossed the field of view. Systems for implementing the method are also disclosed. The invention may be used, in particular, with highway toll booths and for any other application demanding an identification of vehicles.

Claims

1. Method for identifying an object (3) in motion, in particular a vehicle, said object moving inside a predetermined identification zone (2) following a predetermined movement axis (A), which method comprises the steps of:

acquiring periodically images (52) in a predetermined field of view (26), essentially vertical and of a width narrower than its height, said field of view (26) cutting the identification zone (2) at a predetermined angle of intersection and defining an observation plane,

checking nature of image background in the field of view (26), to obtain background reference information in absence of the object (3) in the field of view (26), and

processing the images (52) acquired in combination with the background reference information, to extract therefrom a silhouette of the object (3) having crossed the field of view (26).

2. Method as claimed in claim 1, wherein the processing of the image (52) comprises, at the completion of each acquisition of the image, the following substeps of:

differencing (53) between the acquired image and a reference image, leading to a resultant image,

comparing (55,

     57) the resultant image with a predetermined threshold image, leading, if the resultant image exceeds the threshold image, to a step of storing (58) the resultant image and, in an opposite event, to a step of assigning (56) the resultant image as a new reference image, indicating an absence of the object (3) in the field of view (26), said step of storing (58) being followed by a step of acquiring (59) a new image, the step of differencing (60) and then the step of comparing (61,

     62);

whereby detection of the resultant image below the threshold image leads to a step of processing (63) stored resultant images to extract therefrom the silhouette of the object having crossed the observation plane.

3. Method as claimed in claim 1, wherein the checking of the nature of the image background comprises the substep of lighting a surface (9) of the identification zone (2) included in the field of view (26) and also lighting predetermined parts of the observation plane.

4. Method as claimed in claim 3, wherein the lighting substep is synchronized with the step of periodically acquiring images.

5. Method as claimed in claim 3, wherein the lighting substep is carried out continuously.

6. Method as claimed in claim 3, wherein the lighting substep is carried out periodically with a predetermined frequency, preferably highly relative to a frequency of the step of periodically acquiring images.

7. Method as claimed in claim 1, wherein the step of processing the images acquired comprises a substep of determining (200) specific geometric characteristics of the object (3) to be identified, from the extracted silhouette, followed by a substep of associating the object (3) with a category defined by a predetermined combination of specific geometric characteristics.

8. Method as claimed in claim 7, wherein the step of determining (200) specific geometric characteristics comprises a substep (162) of determining whether the object (3) is constituted from at least two distinct parts joined together by a connecting structure.

9. Method as claimed in claim 8, wherein the step of determining (200) specific geometric characteristics comprises a substep (163) of seeking relative minima of the silhouette of the object (3), corresponding to parts of the object (3) in contact with a surface of the identification zone (2).

10. Method as claimed in claim 1, wherein the step of processing the images acquired comprises a substep of determining height of the object (3) identified, from the extracted silhouette and from predetermined information on localization of the field of view (26) relative to the identification zone (2).

11. Method as claimed in claim 1, which further comprises a step of detecting the presence of the object (3) in a predetermined part of the identification zone (2).

12. Method as claimed in claim 1, wherein the field of view (26) is limited so that only a part of the silhouette of the object (3) is acquired.

13. Method as claimed in claim 2, wherein the step of storing the resultant image comprises a substep of memorizing solely extreme contours of the silhouette of the object (3) under observation.

14. Method as claimed in claim 1, which further comprises the steps of:

detecting a presence (103,

     107) of the object (3) in fixed successive detection planes in the identification zone (2), essentially perpendicular to the movement axis (A), situated on either side of the field of view (26) at predetermined respective distances from the field of view (26), to provide unidimensional spatio-temporal information on movement of the object (3) in the identification zone (2),

wherein the step of processing the images acquired further comprises a substep (300) of determining a length of the object (3) following the movement axis (A) from the extracted silhouette and from spatio-temporal information obtained during the detecting step.

15. Method as claimed in claim 14, wherein the detection planes are arranged within the identification zone (2) so that separations between two fixed successive detection planes are essentially equal.

16. Method as claimed in claim 14, wherein the fixed successive detection planes are situated on either side of the observation plane.

17. Method as claimed in claim 14, wherein the length determining substep comprises a further substep of determining speed of the object (3) crossing the fixed successive detection planes, each detected crossing between two fixed successive detection planes providing a corresponding item of speed information, and comprises a further substep of extrapolating motion of the object (3) to a uniformly accelerated motion.

18. System (1, 20, 90) for identifying an object (3) in motion, in particular a vehicle, said object (3) moving inside a predetermined identification zone (2) following a predetermined axis of movement (A), which system comprises:

means (4) for periodically acquiring images in a predetermined field of view (26), essentially vertical and of a width narrower than its height, said field of view (26) cutting the identification zone (2) at a predetermined angle of intersection and defining an observation plane,

means (5, 6, 9,

     10) for checking nature of an image background in the field of view (26) with an aim of obtaining background reference information, and

means (7) for processing the images acquired in combination with the background reference information, and for extracting therefrom a silhouette of the object (3) having crossed the observation plane.

19. System (1, 20) as claimed in claim 18, wherein the means for checking the nature of the image background comprise means (5, 56, 21, 23, 28, 32) for lighting predetermined parts of the field of view (26).

20. System (1, 20) as claimed in claim 19, wherein the means for checking the nature of the image background further comprise means (9, 10) for reflecting light coming from some of the lighting means (5, 6, 28) towards the image acquiring means (4).

21. System (1, 20) as claimed in claim 20, wherein the identification zone (2) includes a predetermined path (2'"'"'), and further wherein the light-reflecting means (9, 10) comprise bands (9) of reflective material placed on a part of the predetermined path (2'"'"') situated in the field of view (26).

22. System (1, 20) as claimed in claim 21, wherein the light-reflecting means (9, 10) further comprise bands (10) of reflective material placed in the field of view (26) on a predetermined background plane (11, 22) opposite the image acquiring means (4).

23. System (1) as claimed in claim 20, wherein some of the lighting means (5, 6) are situated in immediate proximity to the image acquiring means (4).

24. System (20) as claimed in claim 19, wherein some of the lighting means (5, 6, 21, 23, 28, 32) comprise direct lighting means (21, 23) placed in the field of view (26) on a predetermined background plane (22) opposite the image acquiring means (4) and means (28), situated in proximity to the image acquiring means (4), to light a part of the predetermined path (2'"'"') situated in the field of view (26).

25. System (1, 20) as claimed in claim 19, further comprising means (31) for supplying one of the lighting means (32) with energy.

26. System (1, 20) as claimed in claim 25, wherein the energy supplying means (31) are arranged so that one of the lighting means (32) delivers a periodic light with a predetermined frequency.

27. System (1, 20) as claimed in claim 26, wherein the energy supplying means (31) are arranged so that one of the lighting means (32) is synchronous with the periodically image acquiring means (4).

28. System (1, 20) as claimed in claim 18, wherein the image processing means (7) comprises a central calculating means (30) for receiving image information in digitized form coming from the image acquiring means (4) and also comprises a memory (30c) and means (37) for displaying the extracted silhouettes of the objects (3).

29. System (1, 20) as claimed in claim 28, wherein the image acquiring means (4) comprise a linear camera means (4) for digitizing and delivering to the processing means (7) an image taken in the field of view (26).

30. System (90) as claimed in claim 18, further comprising means (42, 43) for detecting a presence of the object (3) to be identified in fixed planes within the identification zone (2), essentially perpendicular to the axis of movement (A) and situated at predetermined respective distances from the observation plane, and means for supplying unidirectional spatio-temporal information on movement of the object (3) in the identification zone (2), wherein the image processing means (7) is arranged to determine an estimation of a length of the object (3) following the axis of movement (A), from the extracted silhouette and from the spatio-temporal information coming from the detecting means (42, 43) via an interface means (40) between the processing means (7) and the detecting means (42, 43).

31. System (90) as claimed in claim 30, wherein the detecting means (42, 43) comprise a network of optical detection devices placed in the fixed planes, each optical detection device comprising an emitter/receiver detector means (42.1, . . . , 42.N) for emitting an optical beam, situated on one side of a movement path (2'"'"'), and a means (43.1, . . . , 43.N) for reflecting the optical beam towards the detector means (42.1, . . . , 42.N), situated on an opposite side of the movement path (2'"'"'), when the optical beam is not obscured.

32. System (90) as claimed in claim 30, wherein the detecting means are situated on either side of the observation plane.

33. System (90) as claimed in claim 30, wherein the detecting means are equidistant from each other.