Image processing device, cell-cluster recognition apparatus, cell-cluster recognition method, and cell-cluster recognition program for binarizing and segmenting smoothed cell image in which gap between cells in each cell cluster is filled in

1. An image processing device comprising:

a processor comprising hardware, the processor configured to:

smooth a brightness value of a cell image including a plurality of cell clusters each including a plurality of cells so as to generate a smoothed image in which a gap existing between the cells in each of the cell clusters is filled in;

binarize the smoothed image into a background region and a non-background region of each cell cluster, wherein the non-background region includes a plurality of schematic-shape-regions, and the plurality of schematic-shape-regions are spatially independent; and

segment each of the plurality of schematic-shape-regions in the non-background region of the binarized smoothed image into a region for each of the cell clusters.

2. The image processing device according to claim 1 ,

wherein the processor is configured to perform a grayscale tone correction process for increasing a brightness value of a pixel corresponding to the gap within each cell cluster in the cell image prior to being smoothed.

3. The image processing device according to claim 1 ,

wherein the cell image is an image formed by three-dimensionally capturing an image of the plurality of cell clusters.

4. The image processing device according to claim 2 ,

wherein the processor is configured to perform the grayscale tone correction process on at least one of a low-brightness pixel with a brightness lower than or equal to a predetermined low-brightness threshold value and an intermediate-brightness pixel with a brightness higher than the predetermined low-brightness threshold value and lower than or equal to a predetermined high-brightness threshold value.

5. The image processing device according to claim 2 ,

wherein the processor is configured to perform the grayscale tone correction process on the cell image in accordance with a histogram equalization technique.

6. The image processing device according to claim 2 ,

wherein the processor is configured to perform the grayscale tone correction process on the cell image based on a predetermined gamma curve.

7. The image processing device according to claim 1 ,

wherein the processor is configured to set an adaptive predetermined binarization threshold value based on a distribution of pixel values of the cell image and binarize the smoothed image based on the predetermined binarization threshold value.

8. The image processing device according to claim 1 ,

wherein the processor is configured to binarize the smoothed image based on a predetermined binarization threshold value set by a user.

9. The image processing device according to claim 1 ,

wherein the processor is configured to:

calculate a distance value from a boundary with the background region with respect to each pixel within the non-background region;

set an initial region corresponding to each of the cell clusters based on the calculated distance value; and

expand each of the initial regions so as to define a region of each of the cell clusters.

10. The image processing device according to claim 9 ,

wherein the processor is configured to set a region with the distance value larger than a predetermined distance threshold value as the initial region.

11. The image processing device according to claim 1 ,

wherein the processor is configured to convert a brightness value of each pixel to a predetermined grayscale range with respect to the cell image prior to being smoothed.

12. The image processing device according to claim 1 ,

wherein the processor is configured to:

reduce a resolution of the cell image prior to being smoothed based on a predetermined reduction rate; and

increase a resolution of the smoothed image region-segmented for each cell cluster.

13. The image processing device according to claim 1 ,

wherein the processor is configured to:

measure the size of a region of each cell cluster obtained as a result of segmenting the non-background region; and

remove the region of the cell cluster if the measured size thereof is smaller than a predetermined region threshold value.

14. The image processing device according to claim 1 ,

wherein the smoothing becomes stronger with increasing optical magnification when the cell image is captured.

15. A cell-cluster recognition apparatus comprising:

an image acquisition device that is configured to acquire a cell image formed by capturing an image of a plurality of cell clusters each including a plurality of cells; and

an image processing device that includes a processor configured to:

smooth a brightness value of the cell image acquired by the image acquisition device so as to generate a smoothed image in which a gap existing between the cells in each of the cell clusters is filled in;

binarize the smoothed image into a background region and a non-background region of each cell cluster, wherein the non-background region includes a plurality of schematic-shape-regions, and the plurality of schematic-shape-regions are spatially independent; and

segment each of the plurality of schematic-shape-regions in the non-background region of the binarized smoothed image into a region for each of the cell clusters.

16. The cell-cluster recognition apparatus according to claim 15 ,

wherein the processor is configured to perform a grayscale tone correction process for increasing a brightness value of a pixel corresponding to the gap within each cell cluster in the cell image prior to being smoothed.

17. The cell-cluster recognition apparatus according to claim 15 ,

wherein the cell image is an image formed by three-dimensionally capturing an image of the plurality of cell clusters.

18. A cell-cluster recognition method comprising:

smoothing, by an image processor, a brightness value of a cell image including a plurality of cell clusters each including a plurality of cells so as to generate a smoothed image in which a gap existing between the cells in each of the cell clusters is filled in;

binarizing, by the image processor, the smoothed image into a background region and a non-background region of each cell cluster, wherein the non-background region includes a plurality of schematic-shape-regions, and the plurality of schematic-shape-regions are spatially independent; and

segmenting, by the image processor, the non-background region of the binarized smoothed image into a region for each of the cell clusters.

19. A non-transitory computer-readable medium having a cell-cluster recognition program stored thereon, the cell-cluster recognition program causing a computer to execute functions of:

smoothing a brightness value of a cell image including a plurality of cell clusters each including a plurality of cells so as to generate a smoothed image in which a gap existing between the cells in each of the cell clusters is filled in;

binarizing the smoothed image into a background region and a non-background region of each cell cluster, wherein the non-background region includes a plurality of schematic-shape-regions, and the plurality of schematic-shape-regions are spatially independent; and

segmenting each of the plurality of schematic-shape-regions in the non-background region of the binarized smoothed image into a region for each of the cell clusters.