Object recognition system and method

• US 6,069,696 A
• Filed: 06/07/1996
• Issued: 05/30/2000
• Est. Priority Date: 06/08/1995
• Status: Expired due to Term

• Current Assignee: PSC Scanning, Inc. (Datalogic ADC Incorporated)
• Original Assignee: PSC Scanning, Inc. (Datalogic ADC Incorporated)
• Inventors: McMahon, Steven A., Latimer, David L., Turkal, Randy J., McQueen, Alexander M., Cherry, Craig D., Reddersen, Brad R., Schler, Matt D., Rando, Joseph F.
• Primary Examiner(s): HANTIS, KAREN
• Application Number:
  US08/663,324
• Time in Patent Office: 1,453 Days
• Field of Search: 356/402-411, 356/326, 356/328, 356/425, 250/226, 235/467, 235/383, 235/462, 235/462.2-462.31, 235/462.32-462.41, 235/381, 209/587, 186/61
• US Class Current: 356/326
• CPC Class Codes: G01G 19/4144 for controlling weight of goods in commercial establishments, e.g. supermarket, P.O.S. systems
• CPC Class Codes: G01J 3/50 using electric radiation detectors
• CPC Class Codes: G01J 3/501 Colorimeters using spectrally-selective light sources, e.g. LEDs
• CPC Class Codes: G01J 3/502 using a dispersive element, e.g. grating, prism
• CPC Class Codes: G01J 3/51 using colour filters
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06F 2218/00 Aspects of pattern recognition specially adapted for signal processing
• CPC Class Codes: G06V 10/10 Image acquisition document image scanning and transmission H04N1/00; control of digital cameras H04N23/60
• CPC Class Codes: G06V 20/68 Food, e.g. fruit or vegetables

Abstract

An object recognition system comprises a sensing apparatus for collecting light reflected from objects presented at a point-of-sale machine. The sensing apparatus includes a mechanism, such as a holographic disk or diffraction grating, for separating the color components of the light reflected from the object and directing, the color components onto an optical detector such as a two-dimensional imaging array, or a one-dimensional imaging array or single photo-sensitive optical cell used in conjunction with a rotating mirror. A pattern recognizer compares the spectral response, including the locations, amplitudes and widths of energy peaks of the different color components, against premeasured characteristics of known objects in order to classify the object. The weight of the object can be measured with a scale, and the density of the object calculated, with the weight and density being used by the pattern recognizer to further classify the object. In one embodiment, a plurality of narrowband illumination sources are operated in time-sequential manner, each illuminating with a different wavelength band of light, in order to allow separate color measurements. The object recognition system may be integrated in a single unit along with an optical code reader, and may share all or part of the same exit aperture therewith. The object recognition system may include thermal detection or a particle source and secondary emission detection device, either alone or in conjunction with other object recognition means.

Claims

1. An apparatus comprising:

a housing having an inspection aperture;

a light collection system positioned on one side of said inspection aperture so as to receive light reflected from objects located on the other side of said inspection aperture;

a light separator connected to said light collection system, whereby light collected by said light collection system is separated into different color components corresponding to colors of an object to be identified;

an optical detector positioned so as to receive said different color components, said optical detector having an output signal dependent upon intensities of said different color components, said output signal having energy peaks for each of said different color components present in said object to be identified; and

a pattern recognition system connected to said optical detector and responsive to said output signal, wherein said pattern recognition system classifies said object based upon the relative widths of said energy peaks for a plurality of said different color components.

2. The apparatus of claim 1 wherein said light separator comprises a diffraction grating.

3. The apparatus of claim 2 wherein said optical detector comprises a two-dimensional imaging array.

4. The apparatus of claim 2 wherein said optical detector comprises a linear imaging array.

5. The apparatus of claim 4 further comprising a rotating mirror positioned so as to allow said light collection system to build up a two-dimensional image of wavelength versus position by detecting successive images as said rotating mirror rotates.

6. The apparatus of claim 4 wherein said different color components are received by different parts of said one-dimensional imaging array, such that all of said different color components are simultaneously received by said linear imaging array.

7. The apparatus of claim 1 wherein said light separator comprises a hologram.

8. The apparatus of claim 7 wherein said hologram comprises a holographic disk.

9. The apparatus of claim 8 further comprising means for rotating said holographic disk, whereby the color component received by said optical detector changes as said holographic disk is rotated.

10. The apparatus of claim 7 wherein said optical detector comprises a one-dimensional imaging array.

11. The apparatus of claim 1 wherein said light separator comprises a plurality of narrowband illumination sources operating in a time-sequential fashion such that only one of said plurality of narrowband illumination sources illuminates said object at a time.

12. The apparatus of claim 1 further comprising a white light source.

13. The apparatus of claim 1 further comprising an optical code reader sharing a housing and an exit aperture with said light collection system, said light separator and said optical detector.

14. The apparatus of claim 1 wherein said pattern recognition system classifies said object by comparing within a multi-dimensional feature space characteristics of said object with characteristics of known objects, said characteristics of said object comprising at least the relative widths of energy peaks for a plurality of said different color components and the number of energy peaks present for each of said different color components.

15. The apparatus of claim 14 further comprising a database of said known objects connected to said pattern recognition system.

16. The apparatus of claim 1, further comprising a scale for measuring a weight of said objects while said light collection system receives light reflected therefrom, wherein the classification by said pattern recognition system of said object to be recognized is further based upon said weight.

17. The apparatus of claim 1, wherein the classification by said pattern recognition system of said object to be recognized is based at least in part upon variations in intensity within said energy peaks.

18. The apparatus of claim 1, further comprising a white light source for illuminating the object to be recognized, and a lookup correction table accessible to said pattern recognition system comprising a list of correction factors for said different color components to compensate for spectral variations in light output from said white light source.

19. A method comprising the steps of:

collecting light reflected from an object through an inspection aperture;

separating the collected light into different color components;

directing the different color components onto an optical detector;

generating an output signal from said optical detector in response to said different color components;

detecting energy peaks in said output signal for each of said different color components;

measuring widths of said energy peaks; and

classifying said object by comparing at least the relative widths of energy peaks for a plurality of said different color components against premeasured characteristics of known objects.

20. The method of claim 19 wherein said step of separating the collected light into different color components comprises the step of diffracting said collected light such that the direction of travel of each color component of the diffracted light depends upon the wavelength of said color component.

21. The method of claim 20 wherein different color components of said collected light are directed onto different areas of said optical detector.

22. The method of claim 20 wherein said step of directing the different color components onto an optical detector comprises the step of directing the different color components onto a linear imaging array, said method further comprising the step of changing over time the color component directed onto said optical detector by rotating a mirror.

23. The method of claim 19 wherein said step of directing the different color components onto an optical detector comprises the step of directing the different color components onto a linear imaging array, and wherein said step of separating the collected light into different color components comprises the step of sequentially selecting different color bands of light to direct onto said optical detector.

24. The method of claim 23 wherein said step of sequentially selecting different color bands of light to direct onto said optical detector comprises the step of passing said collected light through facets of a holographic disk, each facet of said holographic disk allowing one of said color bands of light to pass through.

25. The method of claim 19 wherein said step of separating the collected light into different color components comprises the step of illuminating said object with light emitted by narrowband illumination sources each differing from the others at least partially in color band of emitted light, only one of said narrowband illumination sources illuminating said object at a time.

26. The method of claim 19, further comprising the step of measuring a weight of said object while collecting light reflected from said object, wherein said step of classifying said object further comprises the step of comparing said measured weight with the premeasured characteristics of known objects.

27. The method of claim 19, further comprising the steps of measuring a weight of said object, measuring a footprint of said object, and calculating a pseudo-density of said object from a ratio of said measured weight to an area of said footprint, wherein said step of classifying said object further comprises the step of comparing said pseudo-density with the premeasured characteristics of known objects.

28. The method of claim 19, further comprising the steps of measuring a weight of said object, measuring a volume of said object, and calculating a density from a ratio of said measured weight to said volume, wherein said step of classifying said object further comprises the step of comparing said density with the premeasured characteristics of known objects.

29. The method of claim 19, wherein said step of classifying said object further comprises the step of comparing variations in intensity within said energy peaks against premeasured characteristics of known objects.

30. The method of claim 19, further comprising the steps of

illuminating said object using a white light source, and

compensating for spectral variations in light output from said white light source by applying a correction factor to said output signal from said optical detector.

31. A system comprising:

a light collection system including at least one lens and aperture, said light collection system collecting light reflected from an object to be identified;

an optical detector;

means connected to said light collection system for separating light collected by said light collection system into different color bands of light corresponding to colors of said object and for directing each color band of light onto said optical detector, said optical detector generating an output signal having an amplitude corresponding to the intensity of light received by said optical detector;

a pattern recognition system connected to said optical detector and responsive to said output signal, said object being classified by said pattern recognition system according to measured widths of energy peaks in said different color bands; and

an optical code recognizer sharing said aperture with said light collection system, said optical code recognizor comprising a decoder for identifying an encoded symbol affixed to said object.

32. The apparatus of claim 31 wherein said optical detector comprises a single photo-sensitive cell.

33. The apparatus of claim 32 further comprising a rotating mirror, whereby the color band received by said photo-sensitive cell changes as said rotating mirror rotates.

34. The apparatus of claim 32 wherein said output signal is indicative of an average color intensity of said object for each of said different color bands.

35. The object recognition system of claim 31 wherein said means for separating light collected by said light collection system into different color bands of light and for directing each color band of light onto said optical detector comprises a diffraction grating positioned such that different color bands of light are directed onto different areas of said optical detector.

36. The object recognition system of claim 31 wherein said means for separating light collected by said light collection system into different color bands of light and for directing each color band of light onto said optical detector comprises a rotating holographic disk having a plurality of facets, each facet directing a different color band of light onto said optical detector.

37. The object recognition system of claim 31 wherein said means for separating light collected by said light collection system into different color bands of light and for directing each color band of light onto said optical detector comprises a plurality of narrowband illumination sources each emitting a different color band of light and operated time-sequentially such that said object is illuminated by only one emitted color band of light at a time.

38. The object recognition system of claim 31 wherein said pattern recognition system comprises a library of known objects defined within a multi-dimensional feature space, said object recognition system classifying said object by comparing said object with said known objects within said multi-dimensional feature space.

39. An apparatus comprising:

an optical code reader, said optical code reader comprising

an optical code reader light collection system including an aperture and a lens,

a photosensitive element positioned so as to receive light gathered by said optical code reader light collection system, and

a decoder connected to said photosensitive element; and

an object recognition system, said object recognition system comprising an object recognition system light collection system sharing said aperture with said optical code reader light collection system.

40. The apparatus of claim 39 wherein said object recognition system further comprises

a light separator connected to said object recognition system light collection system, whereby light collected by said object recognition system light collection system is separated into different color components corresponding to colors of an object to be identified;

an optical detector positioned so as to receive said different color components, said optical detector generating an output signal having energy peaks dependent upon intensities of each of said different color components along a spatial axis as viewed through said aperture; and

a pattern recognition system connected to said optical detector and responsive to said output signal, said object being classified by said pattern recognition system according to the energy peaks in said output signal.

41. The apparatus of claim 39, further comprising a housing containing said optical code reader and said object recognition system.

42. The apparatus of claim 41, wherein said housing comprises a platform for placing items, a first housing portion located beneath said platform, and a second housing portion located next to and above said platform, said first housing portion comprising said aperture and said second housing portion having a second aperture through which light reflected from said items may pass, and

wherein said optical code reader further comprises a second optical code reader light collection system positioned to view said items through said second aperture.

43. The apparatus of claim 39, wherein said decoder is configured to recognize and decode bar codes.

44. An apparatus comprising:

a housing, said housing having a surface atop which items may be placed and a first aperture and second aperture for viewing said items;

a multi-planar optical code reader, said multi-planar optical code reader comprising a first partial field-of-view code reader utilizing said first aperture for viewing said items and a second partial field-of-view code reader utilizing said second aperture for viewing said items, said first partial field-of-view code reader and said second partial field-of-view code reader having at least partially overlapping fields of view; and

an object recognition system sharing at least one of said first aperture and said second aperture with said multi-planar optical code reader.

45. The apparatus of claim 44, wherein said housing comprises a base enclosure portion in which said first aperture is located and an upper housing enclosure portion in which said second aperture is located, said surface atop which items may be placed being located on a top surface of a platform above said base enclosure.

46. The apparatus of claim 44, wherein said object recognition system comprises:

a light collection system;

a light separator for receiving light collected by said light collection system and separating the received light into different color components corresponding to colors of said items placed atop said surface of said housing;

an optical detector positioned so as to receive said different color components, said optical detector having an output signal having energy peaks dependent upon intensities of each of said different color components along a spatial axis as viewed through said at least one of said first aperture and said second aperture shared with said multi-planar optical reader; and

a pattern recognition system connected to said optical detector and responsive to said output signal, said items being classified by said pattern recognition system according to characteristics identified in said output signal, said characteristics including at least said energy peaks.

47. A method comprising the steps of:

receiving, through an aperture, light reflected from an object;

detecting said light using a first photodetector, and generating a first output signal therefrom;

attempting to identify, in response to said first output signal, an encoded symbol;

dividing said light into color components and detecting the different color components using a second photodetector, and generating a second output signal therefrom, said second output signal comprising energy peaks for said different color components; and

attempting to recognize, in response to said second output signal, an object meeting predefined characteristics by evaluating the color of the object as indicated by the color components of said light, and by evaluating the size of the object and its features as indicated by the relative widths and positions of said energy peaks for said different color components.

48. The method of claim 47, wherein said step of detecting said light using a first photodetector and said step of dividing said light into color components and detecting the different color components using a second photodetector are performed simultaneously.

49. An apparatus comprising:

a housing

an object recognizor located within said housing, said object recognizor comprising

a light collection system;

a light separator connected to said light collection system, whereby light collected by said light collection system is separated into different color components corresponding to colors of an object to be identified;

an optical detector positioned so as to receive said different color components, said optical detector having an output signal dependent upon intensities of said different color components; and

a pattern recognition system connected to said optical detector and responsive to said output signal, said object being classified by said pattern recognition system according to widths of energy peaks in said output signal for said different color components; and

an optical code reader located within said housing, said optical reader sharing an aperture with the light collection system of said object recognizor.