Image processing apparatus, image processing method, and computer-readable recording device

1. An image processing apparatus comprising:

a processor comprising hardware, wherein the processor is configured to:

detect an abnormal candidate region, as a candidate for an abnormal portion, from an intraluminal image obtained by capturing an image of an inside of a lumen of a subject;

detect a tubular region from the intraluminal image;

determine whether the abnormal candidate region and the tubular region are overlapped with each other at least in part;

in response to determining that the abnormal candidate region and the tubular region are not overlapped with each other at least in part, determine whether the abnormal candidate region and the tubular region are connected by a connecting region of a color similar to a color of the tubular region;

in response to determining that the abnormal candidate region and the tubular region are not connected by the connecting region of the color similar to the color of the tubular region, determine that the abnormal candidate region is the abnormal portion; and

in response to determining that the abnormal candidate region is the abnormal portion, control a display to display the abnormal portion.

2. The image processing apparatus according to claim 1 ,

wherein the processor is configured to:

calculate inter-region feature data of the connecting region connecting the abnormal candidate region and the tubular region;

determine connectivity between the abnormal candidate region and the tubular region based on the inter-region feature data; and

determine whether the color of the connection region is similar to the color of the tubular region based on the connectivity between the abnormal candidate region and the tubular region.

3. The image processing apparatus according to claim 2 ,

wherein the processor is configured to:

calculate an interpolation line between the abnormal candidate region and the tubular region; and

calculate feature data on the interpolation line as the inter-region feature data.

4. The image processing apparatus according to claim 3 ,

wherein the processor is configured to calculate an edge strength on the interpolation line, that is, a maximum value of changes in pixel values between adjacent pixels or pixels at specified intervals therebetween on the interpolation line as the inter-region feature data.

5. The image processing apparatus according to claim 3 ,

wherein the processor is configured to calculate a color edge strength on the interpolation line, that is, a maximum value of changes in color feature data between adjacent pixels or pixels at specified intervals therebetween on the interpolation line as the inter-region feature data.

6. The image processing apparatus according to claim 3 ,

wherein the processor is configured to calculate an average value of color feature data on the interpolation line as the inter-region feature data.

7. The image processing apparatus according to claim 1 ,

wherein the processor is configured to:

detect a tubular candidate region as a candidate for the tubular region, based on color feature data in the intraluminal image; and

determine whether the tubular candidate region is the tubular region based on continuity of pixel values in a surrounding region of the tubular candidate region.

8. The image processing apparatus according to claim 7 ,

wherein the processor is configured to:

calculate a direction orthogonal to a longitudinal direction of the tubular candidate region in a plane of the intraluminal image;

calculate approximate functions for approximating changes in pixel values on both sides of the tubular candidate region in the orthogonal direction;

calculate a difference between values of the approximate functions in a same coordinate in the orthogonal direction; and

determine whether the tubular candidate region is tubular region based on the difference.

9. The image processing apparatus according to claim 1 ,

wherein the processor is configured to:

detect a tubular candidate region as a candidate for the tubular region, based on color feature data in the intraluminal image; and

determine whether the tubular candidate region is the tubular region based on an outer shape of the tubular candidate region.

10. The image processing apparatus according to claim 9 ,

wherein the processor is configured to:

calculate an area of the tubular candidate region;

calculate a perimeter of the tubular candidate region; and

determine whether the tubular candidate region is the tubular region based on the area and the perimeter.

11. An image processing method comprising:

detecting an abnormal candidate region, as a candidate for an abnormal portion, from an intraluminal image obtained by capturing an image of an inside of an lumen of a subject;

detecting a tubular region from the intraluminal image;

determining whether the abnormal candidate region and the tubular region are overlapped with each other at least in part;

in response to determining that the abnormal candidate region and the tubular region are not overlapped with each other at least in part, determining whether the abnormal candidate region and the tubular region are connected by a connecting region of a color similar to a color of the tubular region;

in response to determining that the abnormal candidate region and the tubular region are not connected by the connecting region of the color similar to the color of the tubular region, determining that the abnormal candidate region is the abnormal portion; and

in response to determining that the abnormal candidate region is the abnormal portion, controlling a display to display the abnormal portion.

12. A computer-readable recording device with an executable program stored thereon, wherein the program instructs a processor to at least perform:

detecting an abnormal candidate region, as a candidate for an abnormal portion, from an intraluminal image obtained by capturing an image of an inside of a lumen of a subject;

detecting a tubular region from the intraluminal image;

determining whether the abnormal candidate region and the tubular region are overlapped with each other at least in part;

in response to determining that the abnormal candidate region and the tubular region are not overlapped with each other at least in part, determining whether the abnormal candidate region and the tubular region are connected by a connecting region of a color similar to a color of the tubular region;

in response to determining that the abnormal candidate region and the tubular region are not connected by the connecting region of the color similar to the color of the tubular region, determining that the abnormal candidate region is the abnormal portion; and

in response to determining that the abnormal candidate region is the abnormal portion, controlling a display to display the abnormal portion.