Stain image color correcting apparatus, method, and system

1. An image processing apparatus for processing a stained sample image obtained by imaging a stained sample, the image processing apparatus comprising:

a processor comprising hardware configured to implement:

a characteristic amount calculator for calculating a first characteristic amount for each pixel constituting the stained sample image; and

a component ratio calculator for, based on the first characteristic amount calculated by the characteristic amount calculator:

calculating a distance of each of the components representing a distance between the target pixel and each of the components in the first characteristic amount;

totalizing the distance of each of the components for all of the components of the target pixel to calculate a distance of all of the components; and

calculating a component ratio of each of the components in the target pixel by calculating a ratio of the distance of each of the components to the distance of all of the components for all of the components of the target pixel; and

a reference value storage medium for storing a reference value of a second characteristic amount for each of the components;

the processor being further configured to implement:

a component correction coefficient calculator for calculating a component correction coefficient representing a correction coefficient for the second characteristic amount of each of the components, based on the second characteristic amount of each of the components and the reference value thereof stored in the reference value storage medium;

a target pixel correction coefficient calculator for calculating, based on the component ratio calculated by the component ratio calculator and the component correction coefficient calculated by the component correction coefficient calculator, a target pixel correction coefficient representing a correction coefficient for the second characteristic amount of the target pixel by multiplying the component ratio and the component correction coefficient together for each of the components of the target pixel to calculate a component-owned correction coefficient representing a correction coefficient owned by each of the components of the target pixel for the second characteristic amount and totalizing the component-owned correction coefficient of all of the components of the target pixel; and

a characteristic amount corrector for correcting the second characteristic amount of the target pixel, based on the target pixel correction coefficient calculated by the target pixel correction coefficient calculator.

2. The image processing apparatus according to claim 1 ,

wherein the component ratio calculator calculates the component ratio of the components, based on the Euclidean distance between a target pixel and each of the components, in a predetermined characteristic amount space of the first characteristic amount in the stained sample image.

3. The image processing apparatus according to claim 1 ,

wherein the component ratio calculator calculates the component ratio of the components, based on the Mahalanobis distance between a target pixel and each of the components, in a predetermined characteristic amount space of the first characteristic amount in the stained sample image.

4. The image processing apparatus according to claim 1 , wherein the processor is further configured to implement a dye amount estimator for estimating a dye amount in each pixel constituting the stained sample image,

wherein the reference value storage medium stores a dye amount reference value for each of the components,

wherein the component correction coefficient calculator calculates, as the component correction coefficient, a correction coefficient for the dye amount of each of the components, based on the dye amount estimated by the dye amount estimator and the dye amount reference value stored in the reference value storage medium,

wherein the target pixel correction coefficient calculator calculates, as the target pixel correction coefficient, a correction coefficient for the dye amount of the target pixel, based on the component ratio calculated by the component ratio calculator and the component correction coefficient calculated by the component correction coefficient calculator, and

wherein the characteristic amount corrector corrects the dye amount of the target pixel, based on the target pixel correction coefficient calculated by the target pixel correction coefficient calculator.

5. The image processing apparatus according to claim 4 ,

wherein the reference value storage medium stores the dye amount of each of the components in an image of a normally stained sample, as the dye amount reference value.

6. The image processing apparatus according claim 4 , wherein the processor is further configured to implement a display image generator for generating a display image, based on the dye amount corrected by the characteristic amount corrector.

7. The image processing apparatus according to claim 4 , wherein the processor is further configured to implement an optical spectrum estimator for estimating an optical spectrum based on a pixel value of the stained sample image,

wherein the dye amount estimator estimates a dye amount in each pixel constituting the stained sample image, based on the optical spectrum estimated by the optical spectrum estimator.

8. The image processing apparatus according to claim 1 ,

wherein the second characteristic amount is a pixel value,

wherein the reference value storage medium stores a reference pixel value of each of the components,

wherein the component correction coefficient calculator calculates, as the component correction coefficient, a correction coefficient for a pixel value of each of the components, based on the pixel value of each of the components and the reference pixel value stored in the reference value storage medium,

wherein the target pixel correction coefficient calculator calculates, as the target pixel correction coefficient, a correction coefficient for the pixel value of the target pixel, based on the component ratio calculated by the component ratio calculator and the component correction coefficient calculated by the component correction coefficient calculator, and

wherein the characteristic amount corrector corrects the pixel value of the target pixel, based on the target pixel correction coefficient calculated by the target pixel correction coefficient calculator.

9. An image processing method of processing a stained sample image obtained by imaging a stained sample, the method comprising the steps of:

calculating a first characteristic amount for each pixel constituting the stained sample image;

performing, based on the first characteristic amount thus calculated, the steps of:

calculating a distance of each of the components representing a distance between the target pixel and each of the components in the first characteristic amount;

totalizing the distance of each of the components for all of the components of the target pixel to calculate a distance of all of the components; and

calculating a component ratio of each of the components in the target pixel by calculating a ratio of the distance of each of the components to the distance of all of the components for all of the components of the target pixel;

calculating a component correction coefficient representing a correction coefficient for a second characteristic amount of each of the components, based on the second characteristic amount of each of the components and a reference value of the second characteristic amount;

calculating, based on the component ratio and the component correction coefficient thus calculated, a target pixel correction coefficient representing a correction coefficient for the second characteristic amount of the target pixel by multiplying the component ratio and the component correction coefficient together for each of the components of the target pixel to calculate a component-owned correction coefficient representing a correction coefficient owned by each of the components of the target pixel for the second characteristic amount and totalizing the component-owned correction coefficient of all of the components of the target pixel; and

correcting the second characteristic amount of the target pixel, based on the target pixel correction coefficient thus calculated.

10. A non-transitory computer-readable medium having stored thereon an image processing program for processing a stained sample image obtained by imaging a stained sample, the program causing a computer to execute the processes of:

calculating a first characteristic amount for each pixel constituting the stained sample image;

performing, based on the first characteristic amount thus calculated, the steps of:

calculating a distance of each of the components representing a distance between the target pixel and each of the components in the first characteristic amount;

totalizing the distance of each of the components for all of the components of the target pixel to calculate a distance of all of the components; and

calculating a component ratio of each of the components in the target pixel by calculating a ratio of the distance of each of the components to the distance of all of the components for all of the components of the target pixel;

calculating a component correction coefficient representing a correction coefficient for a second characteristic amount of each of the components, based on the second characteristic amount of each of the components and a reference value of the second characteristic amount;

calculating, based on the component ratio and the component correction coefficient thus calculated, a target pixel correction coefficient representing a correction coefficient for the second characteristic amount of the target pixel by multiplying the component ratio and the component correction coefficient together for each of the components of the target pixel to calculate a component-owned correction coefficient representing a correction coefficient owned by each of the components of the target pixel for the second characteristic amount and totalizing the component-owned correction coefficient of all of the components of the target pixel; and

correcting the second characteristic amount of the target pixel, based on the target pixel correction coefficient thus calculated.