Automated microscopy system and method for locating and re-locating objects in an image

• US 4,513,438 A
• Filed: 04/15/1982
• Issued: 04/23/1985
• Est. Priority Date: 04/15/1982
• Status: Expired due to Fees

• Current Assignee: Coulter Electronics, Inc.
• Original Assignee: Coulter Electronics, Inc.
• Inventors: Graham, Marshall D., Cook, David D. Jr.
• Primary Examiner(s): Boudreau, Leo H.
• Application Number:
  US06/368,855
• Time in Patent Office: 1,104 Days
• Field of Search: 382/4-6, 382/23, 382/25, 382/26, 382/44, 382/45, 382/48, 358/107, 356/39, 356/40, 356/72, 356/73, 364/413, 364/415, 364/416
• US Class Current: 382/134
• CPC Class Codes: G02B 21/367 providing an output produced by processing a plurality of individual source images, e.g. image tiling, montage, composite images, depth sectioning, image comparison
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06V 10/24 Aligning, centring, orientation detection or correction of the image
• CPC Class Codes: G06V 10/40 Extraction of image or video features
• CPC Class Codes: G06V 10/70 using pattern recognition or machine learning optical pattern recognition or electronic computations therefor G06V10/88
• CPC Class Codes: G06V 20/693 Acquisition

Abstract

A microscopy system for automatically locating and re-locating objects-of-interest in an image which has one or more objects against a background. Each of the objects-of-interest is characterized by a predetermined set of features. In a preferred form, the invention is specifically directed to an automated microscopy system which is adapted for automatically analyzing biological (e.g., blood or cervical) cells on microscope slides. In other forms of the invention, differing types of objects may be identified as being objects-of-interest during a first automated pass and then re-located during a second pass for further analysis.

Claims

1. A microscopy system for automatically locating objects-of-interest in an image having one or more objects against a background, each of said objects-of-interest being characterized by a predetermined set of features, comprising:

object identifier including

i. detecting means for scanning said image and detecting one or more of said objects during said scanning,

ii. feature extraction means for identifying ones of said detected objects having said predetermined set of features as objects-of-interest,

iii. object marker means for generating a marker signal for each of one or more identified object-of-interest, said marker signal having a feature portion representative of said predetermined set of features of said identified object-of-interest, and a pattern portion representative of the relative distance and relative orientation of a plurality of detected objects neighboring said identified object-of-interest.

2. A system according to claim 1 wherein said object identifier is selectively operable, and further comprising:

selectively operable memory means for storing said marker signals generated by said object identifier means,

selectively operable comparison means for comparing a currently generated marker signal with said stored marker signals, means for detecting a substantial match of the pattern portion of said current marker signal and the pattern portion of one of said stored marker signals, and means for generating an object-re-located signal indicative of said match, and

a controller including means operative in a locate mode for operating said object identifier, and said memory means, and including means operative in a re-locate mode for operating said object identifier and said comparison means.

3. A system according to claim 1 wherein said object marker means further includes means for generating coordinate signals for each identified object-of-interest, said coordinate signals being representative of the location of said identified object-of-interest in said image.

4. A system according to claim 3 wherein said object marker means further includes means for generating coordinate signals for each of said neighboring objects, said coordinate signals being representative of the location of said neighboring objects in said image.

5. A system according to claim 1 wherein said object marker means includes means for generating said pattern portion whereby said pattern portion is representative of radial and angular offsets from said identified object-of-interest to each of said neighboring objects.

6. A system according to claim 1 wherein said object marker means includes means for generating said pattern portion whereby said pattern portion is representative of a plurality of orthogonal translational offsets from said identified object-of-interest to each of said neighboring objects.

7. A system according to claim 1 wherein said detecting means includes means for scanning a two-dimensional image.

8. A system according to claim 1 wherein said detecting means includes means for scanning a three-dimensional image.

9. A system according to claim 2 wherein said object marker means further includes means for generating coordinate signals for each identified object-of-interest, said coordinate signals being representative of the location of said identified object-of-interest in said image.

10. A system according to claim 9 wherein said object marker means further includes means for generating coordinate signals for each of said neighboring objects, said coordinate signals being representative of the location of said neighboring objects in said image.

11. A system according to claim 2 wherein said object marker means includes means for generating said pattern portion whereby said pattern portion is representative of radial and angular offsets from said identified object-of-interest to each of said neighboring objects.

12. A system according to claim 2 wherein said object marker means includes means for generating said pattern portion whereby said pattern portion is representative of a plurality of orthogonal translational offsets from said identified object-of-interest to each of said neighboring objects.

13. A system according to claim 2 wherein said detecting means includes means for scanning a two-dimensional image.

14. A system according to claim 2 wherein said detecting means includes means for scanning a three-dimensional image.

15. A system according to claim 1 or 2 wherein said object marker means includes means for identifying neighboring objects for said identified objects-of-interest, wherein said neighboring objects correspond to objects within a determined distance of said object-of-interest.

16. A system according to claim 15 wherein said determined distance is predetermined.

17. A system according to claim 16 wherein said object marker means includes means for identifying neighboring objects for said identified objects-of-interest, wherein said neighboring objects further correspond to objects within a determined angular sector extending from said object-of-interest.

18. A system according to claim 17 wherein said determined angular segment is predetermined.

19. A system according to claim 17 wherein said determined angular segment is adaptively determined.

20. A system according to claim 15 wherein said determined distance is adaptively determined.

21. A system according to claim 20 wherein said object marker means includes means for identifying neighboring objects for said identified objects-of-interest, wherein said neighboring objects further correspond to objects within a determined angular sector extending from said object-of-interest.

22. A system according to claim 21 wherein said determined angular sector is predetermined.

23. A system according to claim 21 wherein said determined angular sector is adaptively determined.

24. A microscopy method for automatically locating objects-of-interest in an image having one or more objects against a background, each of said objects-of-interest being characterized by a predetermined set of features, comprising the steps of

identifying objects by

i. scanning said image and detecting one or more of said objects during said scanning,

ii. identifying ones of said deleted objects having said predetermined set of features as objects-of-interest,

iii. generating a marker signal for each of one or more identified object-of interest, said marker signal having a feature portion representative of said predetermined set of features of said identified obejct-of-interest, and a pattern portion representative of the relative distance and relative orientation of a plurality of detected objects neighboring said identified object-of-interest.

25. A method according to claim 24 wherein said object identifying step is selectively performed, and comprising the further steps of

selectively storing said marker signals generated by said location-defining means,

selectively comparing a currently generated marker signal with said stored marker signals, detecting a substantial match of the pattern portion of said current marker signal and the pattern portion of one of said stored marker signals, and generating an object-re-located signal indicative of said match, and

in a locate mode performing said object identifying step, and said storing step, and in a re-locate mode performing said object identifying step and said comparing step.

26. A method according to claim 24 or 25 wherein said marker generating step further includes the step of generating coordinate signals for each identified object-of-interest, said coordinate signals being representative of the location of said identified object-of-interest in said image.

27. A method according to claim 26 wherein said marker generating step further includes the step of generating coordinate signals for each of said neighboring objects, said coordinate signals being representative of the location of said neighboring objects in said image.

28. A method according to claim 24 or 25 wherein said marker generating step includes the step of generating said pattern portion whereby said pattern portion is representative of radial and angular offsets from said identified object-of-interest to each of said neighboring objects.

29. A method according to claim 24 or 25 wherein said marker generating step includes the step of generating said pattern portion whereby said pattern portion is representative of a plurality of orthogonal translational offsets from said identified object-of-interest to each of said neighboring objects.