Imaging device and morphological feature data display method
Provided is an image classification device that facilitates efficient creation of teacher data and comprehensive evaluation on a basis of knowledge and experience of the user. The image classification device includes: an image input unit that acquires an image; an image display unit that displays the image acquired by the image input unit; an image analysis unit that calculates a feature amount from the acquired image; a feature amount display unit that displays the calculated feature amount; an extraction and sorting condition input unit that specifies an extraction and sorting condition with regard to the feature amount; an extraction and sorting processing unit that performs extraction and sorting processing based on the condition; an extraction and sorting processing result display unit that displays a result; an user classification input unit that allows a user to input a classification destination with regard to the image; and a user classification result display unit that displays an classification input content.
1. An image classification device comprising:
an image input that acquires an image;
an image display that displays the acquired image;
an image analyzer that calculates a morphological feature amount from the image;
a feature amount display that displays the calculated feature amount;
an extraction and sorting condition input that specifies an extraction and sorting condition with regard to the feature amount;
an extraction and sorting processor that performs extraction and sorting processing of the feature amount based on the input condition;
an extraction and sorting processing result display that displays a processing result of the extraction and sorting processor;
a user classification input that allows a user to input a classification destination with regard to the image; and
a user classification result display that displays an input content from the user classification input, wherein
the image is a cross-sectional image of a cell,
the feature amount is a morphological feature amount of a contour of the cell and a feature amount of an internal structure included in an inner region of the contour,
the cell is a platelet, and
the morphological feature amount of the contour is a feature amount indicating roundness of a platelet external form, and the feature amount of the internal structure includes an area ratio of an open small pipe system of the platelet, an expansion degree of the open small pipe system, and an area ratio of a deep stain region of the platelet.
2. The image classification device according to claim 1 , further comprising:
a classification model creator that creates an automatic classifier that classifies the image based on the processing result of the extraction and sorting processor.
3. The image classification device according to claim 2 , further comprising:
an automatic classification result display that displays a classification result of the image classified by the automatic classifier.
4. The image classification device according to claim 3 , wherein
the automatic classification result display highlights the feature amount based on the classification result of the automatic classifier.
5. The image classification device according to claim 3 , wherein
the user classification result display and the automatic classification result display are displayed side by side on a same screen.
6. The image classification device according to claim 3 , further comprising:
a calculation processor that calculates a matching ratio of the classification result of the automatic classifier and the input content input from the user classification input, wherein
the calculated matching ratio is displayed.
7. The image classification device according to claim 3 , wherein
the user classification result display and the automatic classification result display are displayed side by side on the same screen.
8. The image classification device according to claim 2 , wherein
the classification result display highlights the feature amount based on the classification result of the automatic classifier.
9. A display method of an image classification device including an analysis classifier and a display, wherein
the analysis classifier is configured to
calculate a morphological feature amount from an acquired image, and
perform extraction and sorting processing of the feature amount based on an extraction and sorting condition specified with regard to the calculated feature amount, and
the display is configured to
display the acquired image,
display the calculated feature amount,
display a processing result of the extraction and sorting processing of the feature amount, and
display an input content when a user inputs a classification destination with regard to the image, wherein
the image is a cross-sectional image of a cell,
the feature amount is a morphological feature amount of a contour of the cell and a feature amount of an internal structure included in an inner region of the contour,
the cell is a platelet, and
the morphological feature amount of the contour is a feature amount indicating roundness of a platelet external form, and the feature amount of the internal structure includes an area ratio of an open small pipe system of the platelet, an expansion degree of the open small pipe system, and an area ratio of a deep stain region of the platelet.
10. The display method according to claim 9 , wherein
the analysis classifier is configured to
create an automatic classifier that classifies the image based on the processing result of the extraction and sorting processing of the feature amount.
11. The display method according to claim 10 , wherein
the display is configured to
display an automatic classification result of the image classified by the automatic classifier.
12. The display method according to claim 11 , wherein
the display is configured to
highlight the feature amount based on the automatic classification result of the automatic classifier.
13. The display method according to claim 11 , wherein
the display is configured to
display a user classification result and the automatic classification result side by side on the same screen.
14. A non-transitory computer-readable medium comprising instructions operable, when executed by an image classification device comprising an analysis classifier and a display, to:
calculate a morphological feature amount from an acquired image,
perform extraction and sorting processing of the feature amount based on an extraction and sorting condition specified with regard to the calculated feature amount,
display the acquired image,
display the calculated feature amount,
display a processing result of the extraction and sorting processing of the feature amount, and
display an input content when a user inputs a classification destination with regard to the image, wherein
the image is a cross-sectional image of a cell,
the feature amount is a morphological feature amount of a contour of the cell and a feature amount of an internal structure included in an inner region of the contour,
the cell is a platelet, and
the morphological feature amount of the contour is a feature amount indicating roundness of a platelet external form, and the feature amount of the internal structure includes an area ratio of an open small pipe system of the platelet, an expansion degree of the open small pipe system, and an area ratio of a deep stain region of the platelet.
15. The medium according to claim 14 , further comprising instructions operable to:
create an automatic classifier that classifies the image based on the processing result of the extraction and sorting processing of the feature amount.
16. The medium according to claim 15 , further comprising instructions operable to:
display an automatic classification result of the image classified by the automatic classifier.
17. The medium according to claim 16 , further comprising instructions operable to:
highlight the feature amount based on the automatic classification result of the automatic classifier.
18. The medium according to claim 16 , further comprising instructions operable to:
display a user classification result and the automatic classification result side by side on a same screen.