Systems and methods for producing a dynamic classified image

• US 20060056703A1
• Filed: 09/13/2004
• Published: 03/16/2006
• Est. Priority Date: 09/13/2004
• Status: Active Grant

• Current Assignee: Boston Scientific Scimed, Inc. (Boston Scientific Corporation)
• Original Assignee: Boston Scientific Corporation, Scimed Life Systems Incorporated (Boston Scientific Corporation)
• Inventors: Sathyanarayana, Shashidhar
• Granted Patent: US 7,460,716 B2
• Time in Patent Office: Days
• US Class Current: 382/224
• CPC Class Codes: G06F 18/254 of classification results, e.g. of results related to same input data

Abstract

The systems and methods can apply a plurality of different classifiers to a given input image instead of a single classifier. The plurality of classifiers produce a plurality of classified images based on the input image, which are sequentially displayed to a human observer. The sequential display of the classified images produces a dynamic classified image, in which the classification of the input image varies with time depending on which one of the classified image is displayed at a given instant. The dynamic classified image provides dynamic stimuli that encourages the human visual system to excuse occasional classification errors from a minority of the classified images and to retain generally correct classifications from a majority of the classified images.

Claims

1. A method for generating a dynamic classified image, comprising:

applying a plurality of different classifiers to an input image, wherein each of the plurality of classifiers produces a classified image based on the input image; and

sequentially displaying the classified images from the plurality of classifiers.

2. The method of claim 1, wherein the displaying step sequentially displays the classified images at a rate of between 10 to 30 classified images a second.

3. The method of claim 1, wherein the plurality of classifiers classify the input image using the same set of classes.

4. The method of claim 1, wherein the classified images are color-coded.

5. The method of claim 1, further comprising selecting the classifiers from a group consisting of Bayesian classifiers, k-nearest neighbor classifiers, and neural network classifiers.

6. The method of claim 1, wherein supplying step sequentially supplies the plurality of classifiers to the input image.

7. The method of claim 1, wherein the displaying step sequentially displays the classified images in a predetermined order.

8. The method of claim 1, wherein the displaying step sequentially displays the classified images in a random order.

9. The method of claim 1, wherein dynamic classification is applied to only a user-defined portion of the input image.

10. A system for generating a dynamic classified image, comprising:

a plurality of different classifiers configured to process the same input image, wherein each of the plurality of classifiers produces a classified image based on the input image; and

a dynamic classified image processor coupled to the plurality of classifiers, wherein the dynamic classified image processor sequentially outputs the classified images from the plurality of classifiers.

11. The system of claim 10, wherein the dynamic classified image processor is adapted to sequentially display the classified images at a rate of between 10 to 30 classified images a second.

12. The system of claim 10, wherein the plurality of classifiers classify the input image using the same set of classes.

13. The system of claim 10, wherein the classified images are color-coded.

14. The system of claim 10, wherein the classifiers are selectable from a group consisting of Bayesian classifiers, k-nearest neighbor classifiers, and neural network classifiers.

15. The system of claim 10, wherein the dynamic classified image processor sequentially displays the classified images in a predetermined order.

16. The system of claim 10, wherein the dynamic classified image processor sequentially displays the classified images in a random order.

17. The system of claim 10, wherein dynamic classification is applied to only a user-defined portion of the input image.