Image processing device, program, image processing method, computer-readable medium, and image processing system

• US 9,704,017 B2
• Filed: 03/29/2013
• Issued: 07/11/2017
• Est. Priority Date: 08/24/2012
• Status: Active Grant

• Current Assignee: Fujifilm Corporation (Fujifilm Holdings Corporation)
• Original Assignee: Fuji Xerox Pty Limited (Fujifilm Holdings Corporation)
• Inventors: Kato, Noriji, Usuba, Ryoko, Ozaki, Ryota
• Primary Examiner(s): Aynalem, Nathnael
• Application Number:
  US14/401,862
• Publication Number:
  US 20150124082A1
• Time in Patent Office: 1,565 Days
• Field of Search: 382128, 382133, 382134, 348 79, 348135
• CPC Class Codes: G06V 20/693 Acquisition
• CPC Class Codes: H04N 7/18 Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast

Abstract

An image processing device 20 acquires a plurality of captured images obtained by imaging a sample including a target cell at a plurality of different focal distances, sets an object region for detecting the target cell in at least two of the plurality of captured images, and performs determination about whether or not the target cell is included in the object region based on whether or not an image feature quantity based on change in image information in an image in-plane direction and a focal distance direction of an image in the object region in the plurality of captured images satisfies conditions defined for the target cell.

Claims

1. An image processing apparatus comprising:

at least one hardware processor configured to implement;

a setting unit configured to set an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

a calculation unit configured to calculate a Histograms of Oriented Gradients (HOG) feature quantity in each of the plurality of captured images based on a luminance gradient direction and intensity in the object region and to calculate a vector obtained by combining the calculated HOG feature quantities as an image feature quantity; and

a determination unit configured to determine whether the target cell is included in the object region based on whether the image feature quantity satisfies conditions defined for the target cell.

4. The image processing apparatus according to claim 1,

wherein the target cell is a nucleated red blood cell having a nucleus,

the at least one hardware processor is further configured to implement:

an extraction unit configured to extract a nucleus candidate region to be a candidate of the nucleus from the plurality of captured images based on at least one of color and concentration of the nucleus, and

the setting unit is further configured to set the object region based on the nucleus candidate region and a range of possible size of the target cell.

5. The image processing apparatus according to claim 1, wherein the at least one hardware processor configured to implement:

a machine learning unit configured to mechanically learn a first condition for identifying the target cell and other cells based on a plurality of sample images obtained by imaging a positive example and a negative example of the target cell at a plurality of different focal distances.

6. The image processing apparatus according to claim 5,

wherein the machine learning unit is further configured to mechanically learn a second condition for identifying the target cell and a specific cell having a difference in change in image information in at least the focal distance direction from the target cell based on a plurality of sample images obtained by imaging the target cell and the specific cell at a plurality of different focal distances.

10. The image processing apparatus according to claim 6,

wherein the determination unit is further configured to determine that the target cell is included in the object region when it is determined to be a target cell in the object region under both the first condition and the second condition.

11. The image processing apparatus according to claim 1,

wherein the at least one hardware processor configured to implement:

an image selection unit configured to select a captured image most focused on the sample from the plurality of captured images,

wherein the setting unit is further configured to set the object region for a captured image at a focal distance with a predefined range from the captured image selected by the image selection unit among the plurality of captured images.

2. An image processing apparatus comprising:

at least one hardware processor configured to implement;

a setting unit configured to set an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

a calculation unit configured to calculate a Histograms of Oriented Gradients (HOG) feature quantity in each difference image between two captured images adjacent in focal distance among the plurality of captured images based on a luminance gradient direction and intensity in the object region and to calculate a vector obtained by combining the calculated HOG feature quantities as an image feature quantity; and

a determination unit configured to determine whether the target cell is included in the object region based on whether the image feature quantity satisfies conditions defined for the target cell.

7. The image processing apparatus according to claim 2,

wherein the target cell is a nucleated red blood cell having a nucleus,

the at least one hardware processor configured to implement:

an extraction unit configured to extract a nucleus candidate region to be a candidate of the nucleus from the plurality of captured images based on at least one of color and concentration of the nucleus, and

the setting unit is further configured to set the object region based on the nucleus candidate region and a range of possible size of the target cell.

8. The image processing apparatus according to claim 2,

wherein the at least one hardware processor configured to implement:

a machine learning unit configured to mechanically learn a first condition for identifying the target cell and other cells based on a plurality of sample images obtained by imaging a positive example and a negative example of the target cell at a plurality of different focal distances.

9. The image processing apparatus according to claim 8,

wherein the machine learning unit is further configured to mechanically learn a second condition for identifying the target cell and a specific cell having a difference in change in image information in at least the focal distance direction from the target cell based on a plurality of sample images obtained by imaging the target cell and the specific cell at a plurality of different focal distances.

3. An image processing apparatus comprising:

at least one hardware processor configured to implement;

a setting unit configured to set an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

a calculation unit configured to calculate a three-dimensional Histograms of Oriented Gradients (HOG) feature quantity as an image feature quantity based on a three-dimensional luminance gradient direction having an in-plane direction and a focal distance direction of the plurality of captured images in the object region and intensity; and

a determination unit configured to determine whether the target cell is included in the object region based on whether the image feature quantity satisfies conditions defined for the target cell.

12. An image processing method that causes a computer to execute a process comprising:

setting an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

calculating a Histograms of Oriented Gradients (HOG) feature quantity in each of the plurality of captured images based on a luminance gradient direction and intensity in the object region and calculating a vector obtained by combining the calculated HOG feature quantities as an image feature quantity; and

determining whether the target cell is included in the object region based on whether the image feature quantity satisfies conditions defined for the target cell.

13. An image processing method that causes a computer to execute a process comprising:

setting an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

calculating a Histograms of Oriented Gradients (HOG) feature quantity in each difference image between two captured images adjacent in focal distance among the plurality of captured images based on a luminance gradient direction and intensity in the object region and calculating a vector obtained by combining the calculated HOG feature quantities as an image feature quantity; and

determining whether the target cell is included in the object region based on whether the image feature quantity satisfies conditions defined for the target cell.

14. An image processing method that causes a computer to execute a process comprising:

setting an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

calculating a three-dimensional Histograms of Oriented Gradients (HOG) feature quantity as an image feature quantity based on a three-dimensional luminance gradient direction having an in-plane direction and a focal distance direction of the plurality of captured images in the object region and intensity; and

determining whether the target cell is included in the object region based on whether the image feature quantity satisfies conditions defined for the target cell.

15. A non-transitory computer readable medium storing a program causing a computer to execute a process for image processing, the process comprising:

setting an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

calculating a Histograms of Oriented Gradients (HOG) feature quantity in each of the plurality of captured images based on a luminance gradient direction and intensity in the object region and calculating a vector obtained by combining the calculated HOG feature quantities as an image feature quantity; and

determining whether the target cell is included in the object region based on whether or not the image feature quantity satisfies conditions defined for the target cell.

16. A non-transitory computer readable medium storing a program causing a computer to execute a process for image processing, the process comprising:

setting an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

calculating a Histograms of Oriented Gradients (HOG) feature quantity in each difference image between two captured images adjacent in focal distance among the plurality of captured images based on a luminance gradient direction and intensity in the object region and calculating a vector obtained by combining the calculated HOG feature quantities as an image feature quantity; and

determining whether the target cell is included in the object region based on whether the image feature quantity satisfies conditions defined for the target cell.

17. A non-transitory computer readable medium storing a program causing a computer to execute a process for image processing, the process comprising:

setting an object region for detecting a target cell in each of a plurality of captured images obtained by imaging a sample including the target cell at a plurality of different focal distances;

calculating a three-dimensional Histograms of Oriented Gradients (HOG) feature quantity as an image feature quantity based on a three-dimensional luminance gradient direction having an in-plane direction and a focal distance direction of the plurality of captured images in the object region and intensity; and

determining whether the target cell is included in the object region based on whether the image feature quantity satisfies conditions defined for the target cell.