METHOD AND APPARATUS FOR OBJECT CLASSIFICATION

• US 20110116687A1
• Filed: 05/12/2009
• Published: 05/19/2011
• Est. Priority Date: 05/12/2008
• Status: Abandoned Application

• Current Assignee: Qinetiq Limited (QinetiQ Group PLC), Secretary of State for Business, Innovation and Skills
• Original Assignee: Advantage West Midlands, Qinetiq Limited (QinetiQ Group PLC)
• Inventors: McDonald, Gregor John, Levenets, Douglas Alan, Ellis, Jonathan Stanley
• Application Number:
  US12/992,414
• Publication Number:
  US 20110116687A1
• Time in Patent Office: Days
• US Class Current: 382/105
• CPC Class Codes: G06V 20/54 of traffic, e.g. cars on the road, trains or boats
• CPC Class Codes: G06V 20/625 License plates
• CPC Class Codes: G06V 2201/08 Detecting or categorising vehicles
• CPC Class Codes: G06V 30/242 Division of the character sequences into groups prior to recognition; Selection of dictionaries

Abstract

A method of classifying an object, in particular of classifying a vehicle by evaluating the degree of correlation between an image data set corresponding to an image of at least part of the vehicle and each of a plurality of reference data sets each of which corresponds to a reference image of a vehicle comprises the steps of establishing the vehicle'"'"'s aspect and selecting the plurality of reference data sets from a larger group of such data sets on the basis of the vehicle'"'"'s aspect. By selecting from the larger group only those reference data sets which correspond to the vehicle'"'"'s aspect, faster classification is achieved for a given level of processing resources. Alternatively, for a given processing or classification rate, the level of processing resources may be reduced compared to the prior art.

Claims

1-28. -28. (canceled)

29. A method of classifying an object by evaluating the degree of correlation between an image data set corresponding to an image of at least part of the object and each of a plurality of reference data sets each of which corresponds to a reference image of an object, wherein the method comprises the steps of establishing the object'"'"'s aspect and selecting the plurality of reference data sets from a larger group of such data sets on the basis of the object'"'"'s aspect.

30. A method according to claim 29 wherein the object is a vehicle, and the vehicle'"'"'s aspect is established by use of an automatic number-plate recognition system.

31. A method according to claim 29 wherein the object is a vehicle, and the vehicle'"'"'s aspect is established by evaluating the degree of correlation between an image data set corresponding to an image of the vehicle'"'"'s number-plate and each of a plurality of reference data sets each of which corresponds to a reference image of a vehicle number-plate in a respective aspect.

32. A method according to claim 30 further comprising the step of tracking the position of the vehicle'"'"'s number plate as a function of time and determining the vehicle'"'"'s aspect therefrom.

33. A method according to claim 32, comprising capturing a plurality of images of the vehicle and wherein establishing the vehicle'"'"'s aspect comprises selecting an image from the plurality of images in which the vehicle'"'"'s number plate is in a predetermined position within the image.

34. A method according to claim 29 wherein the step of evaluating the degree of correlation between an image data set and a reference data set is carried out by the steps of:

(a) obtaining respective 2D Fourier transform functions of the image data set and of the reference data set;

(b) forming the product of the two 2D Fourier transform functions;

(c) forming the inverse 2D Fourier transform function of said product; and

(d) recording the value of the maximum peak in the inverse 2D Fourier transform.

35. A method according to claim 34 and further comprising the step of summing the values of n highest maxima in the inverse 2D Fourier transform function, where n=2, 3, 4, 5, 6, 7, 8, 9 or 10.

36. A method according to claim 34 wherein the 2D Fourier transforms are processed to produce phase data prior to formation of the product.

37. A method according to claim 29 wherein the reference data sets are generated from wire-framed models of objects.

38. Apparatus for classifying an object, the apparatus comprising processing means for evaluating the degree of correlation between an image data set corresponding to an image of at least part of the object and each of a plurality of reference data sets each of which corresponds to a reference image of an object, wherein the apparatus further comprises means for establishing the object'"'"'s aspect and means for selecting the plurality of reference data sets from a larger group of such data sets on the basis of the object'"'"'s aspect.

39. Apparatus according to claim 38 wherein the object is a vehicle and comprising automatic number-plate recognition apparatus arranged to establish the vehicle'"'"'s aspect.

40. Apparatus according to claim 38 wherein the object is a vehicle and the processing means is arranged to establish the vehicle'"'"'s aspect by evaluating the degree of correlation between an image data set corresponding to an image of the vehicle'"'"'s number-plate and each of a plurality of reference data sets each of which corresponds to a reference image of a vehicle number-plate in a respective aspect.

41. Apparatus according to claim 39 and further comprising means for tracking the vehicle'"'"'s position as a function of time and determining the vehicle'"'"'s aspect therefrom.

42. Apparatus according to claim 41, comprising means to capture a plurality of images of the vehicle and wherein the means to establish the vehicle'"'"'s aspect comprises a means to select an image from the plurality of images in which the vehicle'"'"'s number plate is in a predetermined position within the image.

43. Apparatus according to claim 38 wherein the processing means is arranged to evaluate the degree of correlation between an image data set and a reference data set by the steps of:

(a) obtaining respective 2D Fourier transform functions of the image data set and of the reference data set;

(b) forming the product of the two 2D Fourier transform functions;

(c) forming the inverse 2D Fourier transform function of said product; and

(d) recording the value of the maximum peak in the inverse 2D Fourier transform.

44. Apparatus according to claim 43 comprising a spatial light modulator (SLM) arranged to receive input data corresponding to the product of the two 2D Fourier transform functions and a source of at least partially coherent light, the SLM being arranged to diffract light from the source to produce an optical field corresponding to the inverse 2D Fourier transform function of said product.

45. Apparatus according to claim 43 wherein the processing means is implemented in a FPGA.

46. Apparatus according to claim 45 wherein the processing means is implemented in a FPGA and the FPGA is adapted to form the product of the respective 2D Fourier transform functions of the image data set and of the reference data set by summing the phase data.

47. A method of classifying an object by evaluating the degree of correlation between an image data set corresponding to an image of at least part of the object and each of a plurality of reference data sets each of which corresponds to a reference image of an object, the method comprising:

receiving a series of images;

identifying a set of images in the series of images containing a fiducial marker capable of identifying an object for classification;

selecting an image from the set of images in which the fiducial marker is in a predetermined position; and

classifying the object by evaluating the degree of correlation between image data sets for the object in the selected image with image data sets for reference images determined as appropriate to images with a fiducial marker in the predetermined position.

48. A method according to claim 47, wherein the object is a vehicle, and wherein the fiducial marker is a vehicle'"'"'s number plate, and wherein the object is identified for classification by automatic recognition of the vehicle'"'"'s number plate.

49. A method according to claim 47, wherein the step of evaluating the degree of correlation between an image data set and a reference data set is carried out by the steps of:

(a) obtaining respective 2D Fourier transform functions of the image data set and of the reference data set;

(b) forming the product of the two 2D Fourier transform functions;

(c) forming the inverse 2D Fourier transform function of said product; and

(d) recording the value of the maximum peak in the inverse 2D Fourier transform.

50. A method according to claim 49, wherein the 2D Fourier transforms are processed to produce phase data prior to formation of the product.