Apparatus and method for optimizing detection of objects in computed tomography data

• US 6,272,230 B1
• Filed: 02/11/1998
• Issued: 08/07/2001
• Est. Priority Date: 02/11/1998
• Status: Expired due to Term

• Current Assignee: Analogic Corporation
• Original Assignee: Analogic Corporation
• Inventors: Hiraoglu, Muzaffer, Crawford, Carl R., Simanovsky, Sergey, Bechwati, Ibrahim M.
• Primary Examiner(s): Au, Amelia M.
• Assistant Examiner(s): Dastouri, Mehrdad
• Application Number:
  US09/022,062
• Time in Patent Office: 1,273 Days
• Field of Search: 382/100, 382/128, 382/131, 382/141, 382/228, 378/4, 378/57, 378/901, 345/420, 345/421, 345/424, 250/363.04, 250/367, 250/390.04
• US Class Current: 382/100
• CPC Class Codes: G01N 33/22 Fuels; Explosives
• CPC Class Codes: G01V 5/226 using tomography
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: G06V 2201/05 Recognition of patterns representing particular kinds of hidden objects, e.g. weapons, explosives, drugs

Abstract

A method of and apparatus for detecting objects in computed tomography (CT) data includes the ability to define the types of objects to be detected, and at least one algorithm related to the detection of each type of object. Multiple types of objects can be detected and distinguished from one another. Each type of object exhibits an object detection rate related to the probability of the system detecting the corresponding object type, and a false detection rate related to the false identification of objects, different from the target objects, as the target objects. An overall system detection rate is related to a combination of the object detection rates. Each type of object can also be associated with a unique object false alarm rate, with a overall false detection rate being related to the combination of object false alarm rates. The overall system and/or object detection rate, and/or the false alarm rate and/or the overall false detection rate can be optimized by modifying at least one algorithm so as to adjust at least one of the object detection rates or object false alarm rate.

Claims

1. A method of measuring and adjusting the performance of a system for detecting objects in computed tomography (CT) data for a region, said method comprising:

defining at least two different types of objects for selective detection, wherein at least one of the types of objects is a type of threat object;

defining at least two algorithms respectively related to the detection of the corresponding types of objects so that all of the algorithms form a part of a detection system;

defining (i) for each object type an object detection rate related to the probability of the system detecting the corresponding object type, and (ii) an overall system detection rate related to a combination of the object detection rates; and

modifying at least one algorithm so as to adjust at least one of the object detection rates and adjust the overall system, detection rate;

wherein the object detection rates of the object types are interdependent, the performance of the system is measured by the object detection rates and overall system detection rate, acceptable performance requires each of the object detection rates to be above a first predetermined threshold, and the overall system detection rate to be above a second predetermined threshold, and the overall system detection rate can be adjusted to achieve a predetermined level of performance by modifying at least one algorithm so as to adjust at least one of the object detection rates.

2. The method of claim 1 wherein at least one of the types of objects is a type of explosive.

3. The method of claim 1 wherein the types of objects are respectively different types of threat objects.

4. The method of claim 1 wherein the types of objects are respectively different types of explosives.

5. The method of claim 1 wherein the region includes at least a portion of an interior of a container.

6. The method of claim 5 wherein the container is a piece of baggage.

7. The method of claim 1 wherein the overall system detection rate is an average of the object detection rates.

8. The method of claim 1, wherein each algorithm exhibits an object false alarm rate relativity the probability of the algorithm mistakenly detecting objects as the corresponding object type, and all of the algorithms exhibit an overall system false alarm rate related to a combination of all of the object false alarm rates;

and

further comprising the step of modifying at least one of the algorithms so as to adjust at least one of the object false alarm rates so as to adjust the overall system false alarm rate;

wherein the object false alarm rates are interdependent, the performance of the system is measured by the object false alarm rate, and overall system false alarm rate, and the overall system false alarm rate can be adjusted to achieve a predetermined level of performance by modifying at least one of the algorithms so as to adjust at least one of the object false alarm rates.

9. The method of claim 8 wherein the overall system false alarm rate is a sum of the object false alarm rates.

49. A method according to claim 1, wherein the algorithms are interdependent such that adjustment of one object detection rate affects at least one other object detection rate.

10. A method of measuring and adjusting the performance of a system for detecting objects in computed tomography (CT) data for a region, said method comprising:

defining at least two different types of objects for selective detection wherein at least one of the types of objects is a type of threat object;

defining at least two algorithms respectively related to the detection of the corresponding types of objects so that all of the algorithms form a part of a detection system;

defining for each object type an object false alarm rate related to the probability of the algorithm related to the detection of the corresponding object type mistakenly detecting objects as the corresponding object type, and an overall system false Alan rate related to a combination of the object false alarm rates; and

modifying at least one algorithm so as to adjust at least one of the object false alarm rates so as to adjust the overall system false alarm rate;

wherein the false alarm rates are interdependent, the performance of the system is measured by the object false alarm rates and overall system false alarm rate, and the overall system false alarm rate can be adjusted to achieve a predetermined level of performance by modifying at least one of the algorithms so as to adjust at least one of the object false alarm rates.

11. The method of claim 10 wherein at least one of the types of objects is a type of explosive.

12. The method of claim 10 wherein the types of objects are respectively different types of treat objects.

13. The method of claim 10 wherein the types of objects are respectively different types of explosives.

14. The method of claim 10 wherein the region includes at least a portion of an interior of a container.

15. The method of claim 14 wherein the container is a piece of baggage.

16. The method of claim 10 wherein the overall system false alarm rate is a sum of the object false alarm rates.

50. A method according to claim 10, wherein the algorithms are interdependent such that adjustment of one object detection rate affects at least one other object detection rate.

17. An apparatus for measuring and adjusting the performance of a system for detecting objects in computed tomography (CT) data for a region, said apparatus comprising:

an object selector constructed and arranged so as to define at least two different types of objects for selective detection, wherein at least one of the types of objects is a type of threat object;

at least two object detection algorithmic subsystems respectively related to the detection of the corresponding types of objects so that all of the algorithmic subsystems form a part of a detection system;

detection rate defining subsystem defining (i) for each of type of object an object detection rate related to the probability of the system detecting the corresponding object type, and (ii) an overall system detection rate related to a combination of the object detection rates; and

an object detection algorithmic subsystem modifier constructed and arranged so that at least one object detection algorithmic subsystem can be modified so as to change the corresponding object detection rate of the corresponding object type and adjust the overall system detection rate;

wherein the object detection rates are interdependent, the performance of the system is measured by the object detection rates and overall system detection rate, acceptable performance requires each of the object detection rates to be above a first predetermined threshold, and the overall system detection rate to be above a second predetermined threshold, and the overall system detection rate can be adjusted to achieve a predetermined level of performance by modifying at least one algorithmic subsystem so as to adjust at least one of the object detection rates.

18. The apparatus of claim 17 wherein at least one of the types of objects is a type of explosive.

19. The apparatus of claim 17 wherein the types of objects are respectively different types of threat objects.

20. The apparatus of claim 17 wherein the types of objects are respectively different types of explosives.

21. The apparatus of claim 17 wherein the region includes at least a portion of an interior of a container.

22. The apparatus of claim 21 wherein the container is a piece of baggage.

23. The apparatus of claim 17 wherein the overall system detection rate is an average of the object detection rates.

24. The apparatus of claim 17, wherein each object detection algorithmic subsystem exhibits (a) an object false alarm rate relating to the probability of the object detection algorithmic subsystem mistakenly detecting objects as the corresponding object type, and (b) an overall system false alarm rate related to a combination of the object false alarm rates;

and

the object detection algorithmic subsystem modifier is constructed and arranged so that the at least one object detection algorithmic subsystem can be modified so as to change the corresponding object false alarm rate of at least one object type and to adjust the overall system false alarm rate;

wherein the object false alarm rates are interdependent, the performance of the system is measured by the object false alarm rates and overall system false alarm rate, and the overall system false alarm rate can be adjusted to achieve a predetermined level of performance by modifying at least one of the algorithmic subsystems so as to adjust at least one of the object false alarm rates.

25. The apparatus of claim 24 wherein the overall system false alarm rate is a sum of the object false alarm rates.

51. An apparatus according to claim 17, wherein the algorithms are interdependent such that adjustment of one object detection rate effects at least one other object detection rate.

26. An apparatus for measuring and adjusting the performance of a system for detecting objects in computed tomography (CT) data for a region, said apparatus comprising:

an object selector constructed and arranged so as to define at least two different types of objects for selective detection, wherein at least one of the types of objects is a type of threat object;

at least two object detection algorithmic subsystems respectively related to the detection of each type of object such that all of the algorithmic systems form a part of a detection system, wherein each subsystem defines an object false alarm rate related to the probability of the corresponding object detection algorithmic subsystem mistakenly detecting objects as the corresponding object type, and wherein the object detection algorithmic subsystems define an overall system false alarm rate related to a combination of the object false alarm rates; and

an object detection algorithmic subsystem modifier constructed and arranged so that at least one of the object detection algorithmic subsystems can be modified so as to change the corresponding object false alarm rate of at least one object type, and adjust the overall system false alarm rate;

wherein the object false alarm rates are interdependent, and the performance of the system is measured by the object false alarm rates and overall system false alarm rate, and the overall system false alarm rate can be adjusted to achieve a predetermined level of performance by modifying at least one of the algorithmic subsystems so as to adjust at least one of the object false alarm rates.

27. The apparatus of claim 26 wherein at least one of the types of objects is a type of explosive.

28. The apparatus of claim 26 wherein the types of objects are respectively different types of threat objects.

29. The apparatus of claim 26 wherein the types of objects are respectively different types of explosives.

30. The apparatus of claim 26 wherein the region includes at least a portion of an interior of a container.

31. The apparatus of claim 30 wherein the container is a piece of baggage.

32. The apparatus of claim 26 wherein the overall system false alarm rate is a sum of the object false alarm rates.

52. An apparatus according to claim 26, wherein the algorithms are interdependent such that modification of one object detection rate effects at least one other object detection rate.

33. A computed tomography (CT) scanning object detection system for detecting objects in a region comprising:

an object selector constructed and arranged so as to define at least two different types of objects for selective detection, wherein at least one of the types of objects is a type of threat object;

a computed tomographic data acquisition subsystem constructed and arranged so as to acquire CT data for the region;

a system performance measurement and detection subsystem comprising;

(i) at least two object detection algorithmic subsystems respectively related to the detection of the corresponding types of objects arranged so that all of the algorithmic subsystems form a part of a detection subsystem;

(ii) a detection rate defining subsystem defining, for each of the corresponding types of objects a corresponding object detection rates respectively related to the probability of the CT scanning object detection system detecting in the CT data the corresponding object types, and wherein the object detection algorithmic subsystems exhibit an overall system detection rate related to a combination of the object detection rates; and

is (iii) an object detection algorithmic subsystem modifier constructed and arranged so that at least one object detection algorithmic subsystem can be adjusted so as change the corresponding object detection rate of the respective object type, and adjust the overall system detection rate;

wherein the object detection rates are interdependent, the performance of the system is measured by the object detection rates and overall system detection, acceptable performance requires each of the object detection rates to be above a first predetermined threshold, and the overall system detection rate to be above a second predetermined threshold, and the overall system detection rate can be adjusted to achieve a predetermined level of performance by modifying at least one algorithmic subsystem so as to adjust at least one of the object detection rates.

34. The CT scanning object detection system of claim 33 wherein at least one of the types of objects is a type of explosive.

35. The CT scanning object detection system of claim 33 wherein the types of objects are respectively different types of threat objects.

36. The CT scanning object detection system of claim 33 wherein the types of objects are respectively different types of explosives.

37. The CT scanning object detection system of claim 33 wherein the region includes at least a portion of an interior of a container.

38. The CT scanning object detection system of claim 37 wherein the container is a piece of baggage.

39. The CT scanning object detection system of claim 33 wherein the overall system detection rate is an average of the object detection rates.

40. The CT scanning object detection system of claim 33

wherein at least one object detection algoritlunic subsystem exhibits an object false alarm rate relating to the probability of the object detection algorithmic subsystem mistakenly detecting different objects the corresponding object type, and all of the object detection algorithmic subsystems exhibit an overall system false alarm rate related to a combination of the object false alarm rates; and

the object detection algorithmic subsystem modifier is constructed and arranged so that the at least one object detection algorithmic subsystem can be modified so as to change the corresponding object false alarm rate of at least one object type and adjust the overall system false alarm rate;

wherein the object false alarm rates are interdependent, the performance of the system is measured by the object false alarm rates and overall system false alarm rate, and the overall system false alarm rate can be adjusted to achieve a predetermined level of performance by modifying at least one of the algorithmic subsystems so as to adjust at least one of the object false alarm rates.

41. The CT scanning object detection system of claim 40 wherein the overall system false alarm rate is a sum of the object false alarm rates.

53. A system according to claim 33, wherein the algorithms are interdependent such that modification of one object detection rate effects at least one other object detection rate.

54. A system according to claim 40, wherein the algorithms are interdependent such that modification of one object detection rate effects at least one other object detection rate.

42. A computed tomography (CT) scanning object detection system for detecting objects in a region comprising:

a computed tomographic data acquisition subsystem constructed and arranged so as to acquire CT data for the region;

a system performance measurement and detection subsystem comprising;

(a) a plurality of object detection algorithmic subsystems forming a part of a detection subsystem and defining a like plurality of types of objects selected for detection, and related to the detection of the corresponding types of objects, wherein at least one of the types of objects is a type of threat object,

(b) a false alarm rate defining subsystem defining (i) for each type of object an object false alarm rate related to the probability of the object detection algorithm subsystem mistakenly detecting objects as the corresponding object type in the CT data for the region, and (ii) all overall system false alarm rate related to a combination of the object false alarm rates; and

(c) an object detection algorithmic subsystem modifier constructed and arranged so that any one of the object detection algorithmic subsystems can be modified so as to change the corresponding object false alarm rate of the respective object type, and adjust the overall system false alarm rate;

wherein the object false alarm rates are interdependent, the performance of the system is measured by the object false alarm rates and overall system false alarm rate, and the overall system false alarm rate can be adjusted to achieve a predetermined level of performance by modifying at least one of the algorithmic subsystems so as to adjust at least one of the object false alarm rates.

43. The CT scamming object detection system of claim 42 wherein at least one of the types of objects is a type of explosive.

44. The CT scanning object detection system of claim 42 wherein the objects are respectively different types of threat objects.

45. The CT scanning object detection system of claim 42 wherein the types of objects arc respectively different types of explosives.

46. The CT scanning object detection system of claim 42 wherein the region includes at least a portion of an interior of a container.

47. The CT scamming object detection system of claim 46 wherein the container is a piece of baggage.

48. The CT scanning object detection system of claim 42 wherein the overall system false alarm rate is a sum of the object false alarm rates.