Image processing apparatus and program storage medium

1. An image processing apparatus that processes time-series images that include a plurality of images captured in time series, the image processing apparatus comprising:

an interval setting unit that sets a time-series interval on the basis of a variation between the plurality of images of the time-series images; and

an area classifying unit that determines a category determination criterion for the time-series interval using a feature of an image in the time-series interval, the category determination criterion being common among the images in the time-series interval set by the interval setting unit and classifies each of at least a part of the plurality of images included in the time-series interval into a plurality of category areas using the category determination criterion.

2. The image processing apparatus according to claim 1 , further comprising a variation computing unit that computes a variation between an image of the time-series images and another image located near the image in a time-line, wherein

the interval setting unit sets the time-series interval on the basis of the variation computed by the variation computing unit.

3. The image processing apparatus according to claim 1 , wherein

the area classifying unit comprises:

a feature computing unit that computes the feature of the image included in the time-series interval; and

a distribution estimating unit that estimates a feature distribution of each category area located in the image included in the time-series interval on the basis of a feature distribution that is a distribution of the feature computed by the feature computing unit, and

the area classifying unit performs, in the each of at least a part of the plurality of images included in the time-series interval, classification into the category areas on the basis of the feature distribution of each category area estimated by the distribution estimating unit.

4. The image processing apparatus according to claim 3 , wherein

the area classifying unit comprises an image selecting unit that selects an image from the time-series interval, and

the distribution estimating unit estimates a feature distribution of each category area located in an image in the time-series interval on the basis of a feature distribution of the image selected by the image selecting unit.

5. The image processing apparatus according to claim 3 , wherein

the distribution estimating unit comprises:

a clustering unit that divides the feature distribution that is the distribution of the feature computed by the feature computing unit; and

a category determining unit that determines a category of each cluster obtained by the clustering unit, and

the distribution estimating unit estimates the feature distribution of each category area on the basis of feature data of each cluster of which the category is determined by the category determining unit.

6. The image processing apparatus according to claim 1 , wherein the interval setting unit sets the time-series interval on the basis of a cumulative value that is obtained by accumulating the variations between images in time sequence.

7. The image processing apparatus according to claim 1 , wherein

the interval setting unit sets the time-series intervals such that parts thereof are overlapped in a time-line, and

the area classifying unit integrates classification results for the time-series intervals to perform, in an image included in the overlapped interval, classification into the category areas.

8. The image processing apparatus according to claim 4 , wherein the image selecting unit selects an image on the basis of a variation between images.

9. The image processing apparatus according to claim 8 , wherein the image selecting unit selects an image on the basis of a cumulative value that is obtained by accumulating the variations between images in time sequence.

10. The image processing apparatus according to claim 5 , wherein the category determining unit changes the category determination criterion for each of the time-series intervals.

11. The image processing apparatus according to claim 2 , wherein the variation computing unit computes a variation by comparing pixels or areas that positionally correspond between images.

12. The image processing apparatus according to claim 2 , wherein the variation computing unit computes a variation by comparing amounts of statistics of pixel values.

13. The image processing apparatus according to claim 3 , wherein the feature computing unit divides an image included in the time-series interval into a plurality of areas and computes a feature for each divided area.

14. The image processing apparatus according to claim 1 , wherein the area classifying unit divides an image included in the time-series interval into a plurality of areas and performs category classification for each divided area.

15. The image processing apparatus according to claim 1 , wherein the time-series images are images inside an intra-body canal.

16. The image processing apparatus according to claim 1 , wherein the category areas include at least any one of a mucous membrane, a content, foam, and a lesioned part.

17. A program storage device having stored thereon instructions that causes a computer, which processes time-series images that include a plurality of images captured in time series, to execute:

an interval setting step of setting a time-series interval on the basis of a variation between the plurality of images of the time-series images; and

an area classifying step of determining a category determination criterion for the time-series interval using a feature of an image in the time-series interval, the category determination criterion being common among the images in the time-series interval set by the interval setting step and classifying each of at least a part of the plurality of images included in the time-series interval into a plurality of category areas using the category determination criterion.

18. An image processing method for processing time-series images that include a plurality of images captured in time series, the image processing method comprising steps of:

setting a time-series interval on the basis of a variation between the plurality of images of the time-series images; and

determining a category determination criterion for the time-series interval using a feature of an image in the time-series interval, the category determination criterion being common among the images in the time-series interval set by the interval setting step and classifying each of at least a part of the plurality of images included in the time-series interval into a plurality of category areas using the category determination criterion.

19. An image processing apparatus according to claim 1 , wherein the area classifying unit sets the category determination criterion based on a plurality of images included in the time series interval.

20. The image processing apparatus according to claim 1 , wherein the area classifying unit comprises:

a feature computing unit that computes features of images included in the time-series interval; and

a distribution estimating unit that estimates a feature distribution of each category area located in the images included in the time-series interval on the basis of a feature distribution that is a distribution of the features computed by the feature computing unit, and

the area classifying unit performs, in the each of at least a part of the plurality of images included in the time-series interval, classification into the category areas on the basis of the feature distribution of each category area estimated by the distribution estimating unit.

21. The image processing apparatus according to claim 1 , wherein:

the interval setting unit sets a plurality of time-series intervals on the basis of variations between the plurality of images in the time-series images

the category determining unit changes the category determination criterion among the time-series intervals.