Region segmentation method for particle images and apparatus thereof

Claim 2

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system; a first process of setting a first group of thresholds in density value for a first group of two or more images selected from the red-colored image, green-colored image and blue-colored image to discriminate between a background region and object regions where the object particles exist, and to extract the object regions differing in density from the background region, and producing a first binary image from the first group of two or more images. wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the two or more images of the first group and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds in the two or more images of the first group; a second process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up a single image selected from the first group of images, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the single image; and a third process of performing a logic operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 3

A region segmentation method for particle images according to claim 1, wherein the particle images are images of stained and non-stained urine sediment particles scattered in a flow.

Claim 4

A region segmentation method for particle images according to claim 1, wherein, prior to the first and second process, using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption. density irregularities which are caused by distortion of the image input optical system and which are present in the first group of images are corrected.

Claim 5

A region segmentation method for particle images according to claim 1, wherein the first group of images is the green-colored image and the redcolored image, and the single image is the green-colored image.

Claim 6

A region segmentation method for particle images according to claim 1, wherein at least a portion of the first process and the second process are parallelly executed.

Claim 7

A region segmentation method for particle images according to claim 1, wherein the first process and the second process are parallelly executed.

Claim 8

A region segmentation method for particle images according to claim 1, wherein the logic operation is a logical OR.

Claim 9

A region segmentation method for particle images according to claim 1, wherein either a difference between sums of density values of pixels contained in each of two small areas neighboring each of the pixels, or an absolute value of the difference between the density sums is taken as the quantity representing the magnitude of change of density.

Claim 10

A region segmentation method for particle images according to claim 8, wherein the small areas have a size of two to four pixels arranged onedimensionally in a specified direction in the single image.

Claim 11

A region segmentation method for particle images according to claim 1, wherein either a weighted sum of the density values of pixels in the small area neighboring each of the pixels or an absolute value of the weighted sum is taken as the quantity representing the magnitude of change of density.

Claim 12

A region segmentation method for particle images according to claim 10, wherein the small area has size of two to four pixels in a specified directions in the single image.

Claim 13

A region segmentation method for particle images according to claim 1, wherein either a variance or a standard deviation of distribution of density of pixels in the small areas neighboring each of the pixels is taken as the quantity representing the magnitude of change of density.

Claim 14

A region segmentation method for particle images according to claim 1, wherein the first process comprises a step of generating a density histogram for each of images of the first group of images to determine the low and high thresholds, wherein the low thresholds are lower than a maximum density value of the density histogram, and the high thresholds are higher than the maximum density value of the density histogram.

Claim 15

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system; a first process of correcting density irregularities present in a first group of two or more images selected from the red-colored image, green-colored image and blue-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a second process of setting a first group of thresholds in density value for the first group of two or more images to discriminate between a background region and object regions where the object particles exist. and to extract the object regions differing in density from the background region, and producing a first binary image from the first group of images. wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the two or more images of the first group. and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds in the two or more images of the first group; a third process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up a single image selected from the first group of images, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the single image, the third process being performed in parallel with the second process; and a fourth process of performing a logic operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object.

Claim 16

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system; a first process of correcting density irregularities present in a first group of two or more images selected from the red-colored image, green-colored image and blue-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a second process of setting a first group of thresholds in density value for the first group of two or more images to discriminate between a background region and object regions where the object particles exist. and to extract the object regions differing in density from the background region, and producing a first binary image from the first group of images. wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the two or more images of the first group. and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds in the two or more images of the first group; a third process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up a single image selected from the first group of images, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the single image, the third process being performed in parallel with the second process; and a fourth process of performing a logic operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object.

Claim 17

A region segmentation method for particle images according to claim 14, wherein the particle images are images of stained and non-stained urine sediment particles scattered in a flow.

Claim 18

A region segmentation method for particle images according to claim 14, wherein the first group of images is the green-colored image and the redcolored image, and the single image is the green-colored image.

Claim 19

A region segmentation method for particle images according to claim 14, wherein either a difference between sums of density values of pixels contained in each of two small areas neighboring each of the pixels, or an absolute value of the difference between the density sums is taken as the quantity representing the magnitude of change of density.

Claim 20

A region segmentation method for particle images according to claim 17, wherein the small areas have a size of two to four pixels arranged onedimensionally in a specified direction in the single image.

Claim 21

A region segmentation method for particle images according to claim 20, wherein the first process comprises a step of generating a density histogram for each of images of the first group of images to determine the low and high thresholds, wherein the low thresholds are lower than a maximum density value of the density histogram, and the high thresholds are higher than the maximum density value of the density histogram.

Claim 22

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image, the green-colored image and a blue-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image, green-colored image and blue-colored image; a first binarization means to set a first group of thresholds in density value for a first group of two or more images selected from the red-colored image, green-colored image and blue-colored image to discriminate between a background region and object regions where the object particles exist. and to extract object regions differing in density from the background region, and producing a first binary image from the first group of images. wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the two or more images of the first group, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds in the two or more images of the first group; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up a single image selected from the first group of images, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a single image; and a third binarization means to perform a logic operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 23

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image, the green-colored image and a blue-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image, green-colored image and blue-colored image; a first binarization means to set a first group of thresholds in density value for a first group of two or more images selected from the red-colored image, green-colored image and blue-colored image to discriminate between a background region and object regions where the object particles exist. and to extract object regions differing in density from the background region, and producing a first binary image from the first group of images. wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the two or more images of the first group, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds in the two or more images of the first group; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up a single image selected from the first group of images, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a single image; and a third binarization means to perform a logic operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 24

A region segmentation apparatus for particle images according to claim 20, wherein the particle images are images of stained and non-stained urine sediment particles scattered in a flow.

Claim 25

A region segmentation apparatus for particle images according to claim 20, further comprising a means to correct density irregularities present in the first groupl of images using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system.

Claim 26

A region segmentation apparatus for particle images according to claim 20, wherein the first group of images is the green-colored image and the redcolored image, and the single image is the green-colored image.

Claim 27

A region segmentation apparatus for particle images according to claim 20, wherein at least a portion of an operation of each of the first binarization means and the second binarization means are at least parallelly operated.

Claim 28

A region segmentation apparatus for particle images according to claim 20, wherein the first binarization means and the second binarization means are at least parallelly operated.

Claim 29

A region segmentation apparatus for particle images according to claim 20, wherein the logic operation in the third binarization means is a logical OR.

Claim 30

A region segmentation apparatus for particle images according to claim 20, further comprising a means to determine either a difference between sums of density values of pixels contained in each of two small areas neighboring each of the pixels or an absolute value of the difference between the density sums, wherein either the difference between the density sums or the absolute value of the difference is taken as the quantity representing the magnitude of change of density.

Claim 31

A region segmentation apparatus for particle images according to claim 27, wherein the small areas have a size of two to four pixels arranged one-dimensionally in a specified direction in the single image.

Claim 32

A region segmentation apparatus for particle images according to claim 20, further comprising a means to determine either a weighted sum of the density values of pixels in the small area neighboring each of the pixels or an absolute value of the weighted sum, wherein either the weighted sum or the absolute value of the weighted sum is taken as the quantity representing the magnitude of change of density.

Claim 33

A region segmentation apparatus for particle images according to claim 29, wherein the small area has size of two to four pixels in a specified directions in the group image.

Claim 34

A region segmentation method for particle images according to claim 20, further comprising a means to determine either a variance or a standard deviation of distribution of density of pixels in the small areas neighboring each of the pixels, wherein either the variance or the standard deviation is taken as the quantity representing the magnitude of change of density.

Claim 35

A region segmentation apparatus for particle images according to claim 20, wherein the first binarization means comprises a threshold calculating unit to generate a density histogram for each of images of the first group of images and to determine the low and high thresholds, wherein the low thresholds are lower than a maximum density value of the density histogram, and the high thresholds are higher than the maximum density value of the density histogram.

Claim 36

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image, a green-colored image and a blue-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image, green-colored image and blue-colored image; a means to correct density irregularities present in a first group of two or more images selected from using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption. the density irregularities being caused by distortion of the image input optical system; a first binarization means to set a first group of thresholds in density value for the first group ofimages to discriminate between a background region and object regions where the object particles exist. and to extract object regions differing in density from the background region, and producing a first binary image from the first group of two or more images wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the two or more images of the first group. and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the thresholds in the two or more images of the first group; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up a single image selected from the first group of images, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a second binary image from the single image, the second binarization means being operated in parallel with the first binarization means; and a third binarization means to perform a logic operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 37

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image, a green-colored image and a blue-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image, green-colored image and blue-colored image; a means to correct density irregularities present in a first group of two or more images selected from using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption. the density irregularities being caused by distortion of the image input optical system; a first binarization means to set a first group of thresholds in density value for the first group ofimages to discriminate between a background region and object regions where the object particles exist. and to extract object regions differing in density from the background region, and producing a first binary image from the first group of two or more images wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the two or more images of the first group. and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the thresholds in the two or more images of the first group; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up a single image selected from the first group of images, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a second binary image from the single image, the second binarization means being operated in parallel with the first binarization means; and a third binarization means to perform a logic operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 38

A region segmentation apparatus for particle images according to claim 33, wherein the particle images are images of stained and non-stained urine sediment particles scattered in a flow.

Claim 39

A region segmentation apparatus for particle images according to claim 33, wherein the first group of images is the green-colored image and the redcolored image, and the single image is the green-colored image.

Claim 40

A region segmentation apparatus for particle images according to claim 33, further comprising a means to determine either a difference between sums of density values of pixels contained in each of two small areas neighboring each of the pixels or an absolute value of the difference between the density sums, wherein either the difference between the density sums or the absolute value of the difference is taken as the quantity representing the magnitude of change of density.

Claim 41

A region segmentation apparatus for particle. images according to claim 36, wherein the small areas have a size of two to four pixels arranged one-dimensionally in a specified direction in the single image.

Claim 42

A region segmentation apparatus for particle images according to claim 33, wherein the first binarization means comprises a threshold calculating unit to generate a density histogram for each of images of the first group of images and to determine the low and high thresholds, wherein the low thresholds are lower than a maximum density value of the density histogram, and the high thresholds are higher than the maximum density value of the density histogram.

Claim 43

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system: a first process of setting a first group of thresholds in density value for two images selected from the red-colored image, green-colored image and blue-colored image to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the two images, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for each of the two images, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a second process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up single image selected from the two images, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the single image; and a third process of performing a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 44

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system: a first process of setting a first group of thresholds in density value for two images selected from the red-colored image, green-colored image and blue-colored image to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the two images, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for each of the two images, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a second process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up single image selected from the two images, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the single image; and a third process of performing a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 45

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system; a first process of correcting density irregularities present in two images selected from the red-colored image, green-colored image and blue-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a second process of setting a first group of thresholds in density value for two images to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the two images, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for each of the two images, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a third process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up single image selected from the two images, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the single image, the third process being performed in parallel with the second process; and a fourth process of performing a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 46

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system; a first process of correcting density irregularities present in two images selected from the red-colored image, green-colored image and blue-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a second process of setting a first group of thresholds in density value for two images to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the two images, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for each of the two images, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a third process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up single image selected from the two images, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the single image, the third process being performed in parallel with the second process; and a fourth process of performing a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 47

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image, a green-colored image and a blue-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image, green-colored image and blue-colored image; a first binarization means to set a first group of thresholds in density value for two images selected from the red-colored image, green-colored image and blue-colored image to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the two images, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for each of the two images, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up single image selected from the two images, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a second binary image from the single image; and a third binarization means to perform a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 48

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image, a green-colored image and a blue-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image, green-colored image and blue-colored image; a first binarization means to set a first group of thresholds in density value for two images selected from the red-colored image, green-colored image and blue-colored image to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the two images, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for each of the two images, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up single image selected from the two images, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a second binary image from the single image; and a third binarization means to perform a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 49

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image, a green-colored image and a blue-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image, green-colored image and blue-colored image; a means to correct density irregularities present in two images selected from the red-colored image, green-colored image and blue-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a first binarization means to set a first group of thresholds in density value for the two images to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the two images, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for each of the two images, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up single image selected from the two images, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a second binary image from the single image, the second binarization means being operated in parallel with the first binarization means; and a third binarization means to perform a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 50

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image, a green-colored image and a blue-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image, green-colored image and blue-colored image; a means to correct density irregularities present in two images selected from the red-colored image, green-colored image and blue-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a first binarization means to set a first group of thresholds in density value for the two images to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the two images, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for each of the two images, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up single image selected from the two images, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a second binary image from the single image, the second binarization means being operated in parallel with the first binarization means; and a third binarization means to perform a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 51

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system; a first process of correcting density irregularities present in the red-colored image and the green-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a second process of setting a first group of thresholds in density value for the red-colored image and green-colored image to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the red-colored image and green-colored image, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the red-colored image and green-colored image, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a third process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up the green-colored image, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the green-colored image, the third process being performed in parallel with the second process; and a fourth process of performing a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 52

A region segmentation method for particle images comprising: photographing and storing as static images a red component image (red-colored image), a green component image (green-colored image) and a blue component image (blue-colored image) of stained and non-stained object particles supplied through an image input optical system; a first process of correcting density irregularities present in the red-colored image and the green-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a second process of setting a first group of thresholds in density value for the red-colored image and green-colored image to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the red-colored image and green-colored image, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the red-colored image and green-colored image, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a third process of calculating for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up the green-colored image, comparing a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, setting a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and producing a second binary image from the green-colored image, the third process being performed in parallel with the second process; and a fourth process of performing a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 53

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image and a green-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image and the green-colored image; a means to correct density irregularities present in the red-colored image and the green-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a first binarization means to set a first group of thresholds in density value for the red-colored image and green-colored image to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the red-colored image and green-colored image, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the red-colored image and green-colored image, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up the green-colored image, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a second binary image from the green-colored image, the second binarization means being operated in parallel with the first binarization means; and a third binarization means to perform a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.

Claim 54

A region segmentation apparatus for particle images comprising: an image input optical system for inputting particle images; a means to generate a red-colored image and a green-colored image of stained and non-stained object particles as static images; a memory means to store image data of the red-colored image and the green-colored image; a means to correct density irregularities present in the red-colored image and the green-colored image using a white image photographed when an object has a uniform and infinitely small light absorption and a black image photographed when an object has an infinitely large light absorption, the density irregularities being caused by distortion of the image input optical system; a first binarization means to set a first group of thresholds in density value for the red-colored image and green-colored image to discriminate between a background region and object regions where the object particles exist, and to extract object regions differing in density from the background region, and producing a first binary image from the red-colored image and green-colored image, wherein the first group of thresholds are low thresholds and high thresholds in density value and the first group of thresholds are set for the red-colored image and green-colored image, and density values of pixels of the extracted object regions are lower than the low thresholds or higher than the high thresholds; a second binarization means to calculate for each pixel a quantity representing the magnitude of change of density in an area neighboring each of the pixels making up the green-colored image, to compare a change of density in the background region with the quantity representing the magnitude of change of density calculated for each pixel, to set a second group of thresholds for extracting object regions whose change of density is larger than that of the background region, and to produce a second binary image from the green-colored image, the second binarization means being operated in parallel with the first binarization means; and a third binarization means to perform a logical OR operation on the first binary image and the second binary image to produce a third binary image showing the background region and the object regions.