JP5449444B2 - Colony counting device, colony counting method, and program - Google Patents

Description

  The present invention relates to a colony counting device, a colony counting method, and a program.

  At present, awareness of food hygiene is increasing. For this reason, strict food hygiene inspections are carried out on domestic and foreign foods. In food hygiene inspection, for example, after culturing a sample collected from food in a medium, the number of colonies of microorganisms such as coliforms generated in the sample is counted, and whether the number of generated colonies is less than a predetermined value. Inspect.

  Currently, a colony counting device capable of automatically counting the number of colonies has been developed. For example, in Patent Document 1, a connection region corresponding to a colony is obtained based on image data obtained by imaging a culture medium in which a colony is generated, and a ratio between a long side and a short side of a rectangle circumscribing the connection region, etc. A colony counting device that counts the number of colonies in consideration of the above is disclosed.

  Such a colony counting apparatus binarizes an image obtained by imaging a culture medium with a threshold value higher than the luminance value of a pixel representing the colony, and performs a labeling process on the binarized image. Count the number of connected components corresponding to the colony. The threshold used for binarization can be set according to the color density of the colony.

Japanese Patent No. 2791303

  However, the color density of the colonies generated in the medium is not uniform, and colonies having various color densities exist simultaneously in the medium. In particular, the coliform group has many types of bacteria, and the metabolic activity varies depending on the bacteria. For example, Escherichia coli group degrades the enzyme substrate X-GAL (5-Bromo-4-Chloro-3-indolyl-β-D-galactopyranoside), and the blue pigment 5,5′-dibromo-4,4′-dichloro-indigo. The concentration varies depending on the bacterial species. In particular, in the case of a film or sheet-like medium, the concentration difference is remarkable, and therefore the number of colonies is very difficult to measure. In S. aureus, similar pigment production has been confirmed by the degradation of X-phosphate.

  Here, for example, a region between a plurality of dark-colored colonies existing at a short distance may have a darker color than a light-colored colony. In such a case, for example, if the threshold value is increased in order to detect light-colored colonies, a plurality of dark-colored colonies existing at a short distance are detected as one colony, and the number of colonies can be accurately counted. Can not. Therefore, a colony counting device capable of accurately counting the number of colonies having various color densities is desired.

  The present invention has been made in view of such circumstances, and provides a colony counting device, a colony counting method, and a program capable of accurately counting the number of colonies having various color densities. The purpose is to do.

In order to achieve the above object, the colony counting device according to the first aspect of the present invention comprises:
A grayscale image acquisition means for acquiring a grayscale image showing the state of the culture medium after culturing the microorganisms contained in the specimen;
Binarized image generating means for generating a binarized image by binarizing the grayscale image acquired by the grayscale image acquiring means by the threshold value every time a threshold is supplied;
Each time the binarized image is generated by the binarized image generating means, a labeling process is performed on the binarized image, thereby connecting a group of pixels indicating one colony from the binarized image. Connected component extraction means for extracting the components;
Each time the connected component is extracted by the connected component extraction means, the storage device stores connected component information indicating a connected component that does not include the connected component indicated by the connected component information stored in the storage device. Connected component information adding means for further storing
Threshold supply means for increasing the threshold value by a predetermined value and supplying it to the binarized image generating means each time processing of the connected component information adding means is completed until the threshold value reaches the upper limit value. When,
Colony counting means for counting the number of colonies of the microorganism based on the connected component information stored in the storage device,
It is characterized by that.

  You may further provide the threshold range determination means which determines the said lower limit and the said upper limit based on distribution of the luminance value of the pixel which comprises the said gray scale image.

  The threshold range determination means identifies a luminance value having the highest frequency in the luminance range of the maximum mountain-shaped frequency distribution included in the luminance histogram representing the frequency for each luminance value of the pixels constituting the grayscale image. The highest luminance value may be determined as the upper limit value among the luminance values that are lower than the specified luminance value and have a frequency equal to or less than a predetermined ratio with respect to the frequency of the specified luminance value.

  The threshold range determining unit may specify a minimum luminance value of luminance values of pixels constituting the gray scale image, and determine the specified luminance value as the lower limit value.

Color image acquisition means for acquiring a color image obtained by imaging the culture medium;
A hue range specifying means for specifying a hue range of the colony based on a hue distribution of pixels constituting the color image acquired by the color image acquiring means;
Among the pixels constituting the image obtained by converting the color image acquired by the color image acquisition unit into a gray scale, the luminance value of the pixel not included in the hue range specified by the hue range specifying unit is determined from the upper limit value. And a gray scale image generating unit that generates a gray scale image acquired by setting a higher luminance value as a gray scale image to be acquired by the gray scale image acquiring unit.

  The hue range specifying means obtains a hue range of a mountain-shaped frequency distribution that exists after excluding the hue range of the maximum mountain-shaped frequency distribution from a hue histogram representing the frequency for each hue of pixels constituting the color image. You may specify as the hue range of the said colony.

Color image acquisition means for acquiring a color image obtained by imaging the culture medium;
Grayscale image generation for generating a grayscale image obtained by applying a Gaussian filter to a grayscale image obtained by converting the color image acquired by the color image acquisition unit to the grayscale image acquisition unit Means may be further provided.

In order to achieve the above object, the colony counting method according to the second aspect of the present invention comprises:
A grayscale image acquisition step of acquiring a grayscale image showing the state of the culture medium after culturing the microorganisms contained in the specimen;
A binarized image generating step of generating a binarized image by binarizing the grayscale image acquired by the grayscale image acquiring step by the threshold value every time a threshold is supplied;
Each time the binarized image is generated by the binarized image generating step, a labeling process is performed on the binarized image, thereby connecting a group of pixels indicating one colony from the binarized image. A connected component extraction step for extracting components;
Each time the connected component is extracted by the connected component extracting step, connected storage information indicating connected components that do not include the connected components indicated by the connected component information stored in the storage device among the connected components is stored in the storage device. Connected component information adding step to be further stored in
A threshold supply step of increasing the threshold value by a predetermined value and executing the binarized image generation step each time the processing of the connected component information adding step is completed until the threshold value reaches the upper limit value. When,
A colony counting step for counting the number of colonies of the microorganism based on the connected component information stored in the storage device,
It is characterized by that.

In order to achieve the above object, a program according to the third aspect of the present invention provides:
Computer
A grayscale image acquisition means for acquiring a grayscale image showing the state of the culture medium after culturing the microorganisms contained in the specimen;
A binarized image generating unit that generates a binarized image by binarizing the grayscale image acquired by the grayscale image acquiring unit by the threshold every time a threshold is supplied;
Each time the binarized image is generated by the binarized image generating means, a labeling process is performed on the binarized image, thereby connecting a group of pixels indicating one colony from the binarized image. Connected component extraction means for extracting components;
Each time the connected component is extracted by the connected component extraction means, the storage device stores connected component information indicating a connected component that does not include the connected component indicated by the connected component information stored in the storage device. Connected component information adding means for further storing
Threshold supply means for increasing the threshold value by a predetermined value and supplying it to the binarized image generating means each time processing of the connected component information adding means is completed until the threshold value reaches the upper limit value. ,
Based on the connected component information stored in the storage device, function as colony counting means for counting the number of colonies of the microorganism,
It is characterized by that.

  According to the present invention, it is possible to accurately count the number of colonies having various color densities.

It is a block diagram which shows the hardware constitutions of the colony counting device which concerns on embodiment of this invention. It is a figure which shows the exclusive sheet | seat with which the some film type culture medium was mounted | worn. It is a figure which shows the mode of the culture medium which the colony of microorganisms generate | occur | produced. It is a figure which shows a binarized image. It is a figure for demonstrating a labeling process. It is a figure which shows the rectangle circumscribing a connection component. It is a figure which shows the functional block of the colony counting device which concerns on embodiment of this invention. It is a flowchart which shows the colony counting process which the colony counting device which concerns on embodiment of this invention performs. It is a figure which shows the 5 * 5 operator used for a Gaussian filter. (A) is a figure which shows the connected component detected when the Gaussian filtering process is not performed. (B) is a figure which shows the connected component detected at the time of Gaussian filtering process execution. It is a flowchart which shows the threshold value range determination process shown in FIG. It is a figure which shows the histogram which shows frequency for every luminance value. It is a flowchart which shows the hue band pass filtering process shown in FIG. It is a figure which shows the histogram which shows frequency for every hue. It is a flowchart which shows the connected component count process shown in FIG. (A) is a figure which shows a gray scale image. (B) is a figure which shows the connected component detected when a threshold value is set to the minimum value. (A) is a figure which shows the connected component detected when a threshold value is increased 1 unit from the lowest value. (B) is a figure which shows the connected component detected when the threshold value after 1 unit increase is set. (A) is a figure which shows the connected component detected when a threshold value is further increased by 1 unit. (B) is a figure which shows the connected component detected when the threshold value after 2 unit increase is set. It is a figure which shows the screen which a colony counting device presents. It is a figure which shows the base material sheet which the whole surface is a culture medium.

  Hereinafter, a colony counting device according to an embodiment of the present invention will be described with reference to the drawings.

(Embodiment)
First, a hardware configuration of the colony counting device 100 according to the present embodiment will be described with reference to FIG. The colony counting device to which the present invention is applied is not limited to the colony counting device 100 shown in FIG. For example, the present invention may be applied to a colony counting device in which various devices are incorporated in the colony counting device 100, or the present invention is applied to a colony counting device in which various devices are appropriately excluded from the colony counting device 100. May be.

  First, the physical configuration of the colony counting apparatus 100 will be described with reference to FIG. As shown in FIG. 1, a colony counting apparatus 100 includes a CPU (Central Processing Unit) 101, a ROM (Read Only Memory) 102, a RAM (Random Access Memory) 103, a storage device 104, an input device 105, a communication device 106, and an imaging control. Unit 107, image storage unit 109, display control unit 110, and sound processing unit 112. These components included in the colony counting apparatus 100 are connected to each other by a bus 120. Note that an imaging device 108 is connected to the imaging control unit 107. In addition, a display device 111 is connected to the display control unit 110. An audio output device 113 is connected to the audio processing unit 112.

  The CPU 101 controls the overall operation of the colony counting device 100. Note that the CPU 101 operates according to a program stored in the ROM 102 and uses the RAM 103 as a work area.

  The ROM 102 stores a program and data for controlling the entire operation of the colony counting apparatus 100.

  The RAM 103 functions as a work area for the CPU 101. That is, the CPU 101 temporarily writes programs and data in the RAM 103 and refers to these programs and data as appropriate.

  Therefore, the CPU 101, the ROM 102, and the RAM 103 constitute a control unit. The control unit controls each unit connected to the bus 120 to execute a colony counting process and the like.

  The storage device 104 stores captured images and parameters. The storage device 104 may be a built-in storage device or a removable storage device. The storage device 104 includes, for example, a hard disk device, a memory card slot including a memory card, a CD-ROM drive including a CD-ROM (Compact Disk Read Only Memory), and the like.

  The input device 105 is a device that receives an operation input from an operator or the like. The input device 105 includes, for example, a mouse, a keyboard, a touch screen, a controller, buttons, and the like.

  Under the control of the CPU 101, the communication device 106 establishes communication with a device outside the colony counting device 100 by wireless communication or wired communication, and receives and transmits data. For example, the communication device 106 receives a captured image captured by the imaging device 108 or the like from a database server connected to the colony counting device 100 via an electric communication network such as the Internet. The communication device 106 includes a NIC (Network Interface Card), a USB (Universal Serial Bus) port, and the like.

  The imaging control unit 107 controls the imaging device 108 connected to the imaging control unit 107 under the control of the CPU 101. For example, the imaging control unit 107 controls the operation of the imaging device 108 by transmitting a control signal indicating an imaging start instruction, an imaging end instruction, or the like to the imaging device 108. In addition, the imaging control unit 107 sets imaging conditions in the imaging device 108 by transmitting imaging settings such as resolution, exposure time, and brightness to the imaging device 108, for example. Then, the imaging control unit 107 receives the captured image acquired by the imaging device 108 from the imaging device 108.

  The imaging device 108 images a subject and generates a captured image indicating the subject. The imaging device 108 images a subject in accordance with a control signal supplied from the imaging control unit 107. The imaging device 108 transmits the generated captured image to the imaging control unit 107. The captured image is, for example, a color image in which each pixel is represented by a luminance value of R (Red), a luminance value of G (Green), and a luminance value of B (Blue). However, the captured image may be a monochrome image in which each pixel is represented by one luminance value. The captured image is represented by digital data, for example. The aspect ratio and resolution of the captured image can be adjusted as appropriate. The imaging device 108 includes a digital camera, a digital video camera, a flat bed scanner, a dedicated reader, and the like.

  In the present embodiment, the subject imaged by the imaging device 108 is the dedicated sheet 200 on which a plurality of culture media after culturing the microorganisms contained in the specimen are mounted. Here, the specimen is collected from a food to be examined. As shown in FIG. 2, for example, a plurality of film-type substrate sheets 210 are mounted on the dedicated sheet 200 in a matrix. FIG. 2 shows an example in which 3 × 2 = 6 substrate sheets 210 are attached to the dedicated sheet 200. The imaging device 108 images this dedicated sheet 200 and generates a captured image. The captured image is, for example, 300 dpi, 2472 × 3496 Pixels, 2.7 Mbytes. However, the resolution and the like are not limited to this example.

  FIG. 3 shows a state in which microbial colonies 250 are generated in the culture layer 230 included in the base sheet 210. In addition, identification information 220 relating to the culture layer 230 is described in the base sheet 210. The identification information 220 is, for example, culture conditions or dates. The microorganism is a fungus such as E. coli. Microorganism colonies 250 are generated in the culture layer 230 by culturing the microorganisms.

  In the present embodiment, a description will be given assuming that a region 240 including the culture layer 230 is extracted from a color captured image obtained by imaging the entire dedicated sheet 200. However, the extracted area is not limited to this example. For example, the extracted region may be a partial region of the culture layer 230. Hereinafter, an image showing the region 240 extracted from a color captured image obtained by imaging the entire dedicated sheet 200 is simply referred to as a color image. The culture layer 230 is appropriately referred to as a medium, and the base sheet 210 is appropriately referred to as a mount.

  Here, since the color of the culture layer 230 and the color of the colony 250 are different, the region representing the culture layer 230 and the region representing the colony 250 can be distinguished on the color image. In this embodiment, the color of the colony 250 is a color close to dark blue-green, the color of the culture layer 230 is a color close to light skin color, and the color of the base sheet 210 is a color close to bright white And In the grayscale image generated when the color image is grayscaled, for example, a pixel having a luminance value lower than the first threshold value lower than the luminance value of the pixel representing the culture layer 230 is a pixel representing the colony 250. A pixel having a luminance value higher than the second threshold value higher than the luminance value of the pixel representing the culture layer 230 is a pixel representing the base sheet 210, and has a luminance higher than the first threshold value and lower than the second threshold value. The value pixel is the color of the culture layer 230. Hereinafter, the “first threshold value” is simply referred to as “threshold value” as appropriate. Details of a method for detecting a pixel representing the colony 250 and a pixel representing the culture layer 230 on the color image will be described later.

  Even without the base sheet 210 (mounting paper), the binarization processing described later can be performed. For example, when the entire surface of the base sheet 210 is the culture layer 230, the pixel representing the colony 250 and the pixel representing the culture layer 230 can be distinguished only by the first threshold value. On the other hand, when the base sheet 210 or the culture layer 230 is provided with a ruled line that forms a lattice of a predetermined size, a threshold value is required to distinguish the pixel representing the ruled line from the pixel representing the colony 250 or the like. Become. This grid is typically square. The size of the lattice indicates a scale used when determining the density of the colony 250.

  In addition, although this embodiment demonstrates the example which employ | adopts the culture layer 230 which is a film thru | or sheet-like dry culture medium, you may employ | adopt the agar medium with which the petri dish was filled. When the culture layer 230 that is a film or sheet-like dry medium is employed, it is preferable to employ a scanner or the like as the imaging device 108. On the other hand, when an agar medium filled in a petri dish is employed, a digital camera or the like is preferably employed as the imaging device 108.

  The image storage unit 109 stores a captured image obtained by imaging the dedicated sheet 200 including the culture layer 230 to be detected by the colony 250. The image storage unit 109 stores, for example, a captured image acquired from the imaging device 108, a captured image stored in the storage device 104, a captured image acquired from an external device by the communication device 106, or the like. The image storage unit 109 may store a color image extracted from the captured image instead of the captured image. The image storage unit 109 is configured by, for example, a flash memory.

  The display control unit 110 controls the display device 111 under the control of the CPU 101. For example, the display control unit 110 generates an image signal representing an image to be displayed on the display device 111 and supplies the generated image signal to the display device 111.

  The display device 111 displays an image based on the image signal supplied from the display control unit 110. The display device 111 is configured by a display device such as an LCD (Liquid Crystal Display) or an organic EL (Electro Luminescence). The display device 111 may be a touch screen. In this case, the touch screen functions not only as a display device but also as an input device 105.

  The audio processing unit 112 converts the digital audio signal supplied from the CPU 101 into an analog audio signal by a D / A (Digital / Analog) converter (not shown) and supplies the analog audio signal to the audio output unit 113. The audio processing unit 112 may include a DSP (Digital Signal Processor) that performs various processes on the digital audio signal supplied from the CPU 101.

  The audio output device 113 converts the analog audio signal supplied from the audio processing unit 112 into audio and outputs the audio. The audio output device 113 is configured by a speaker, for example. The audio output device 113 may include an amplifying device that amplifies a digital audio signal or an analog audio signal.

  Next, with reference to FIGS. 4-6, the labeling process performed in the connected component count process which the colony counting device 100 which concerns on this embodiment performs is demonstrated. In addition to the labeling process described below, various labeling processes can be employed in the present invention. In order to facilitate understanding, the grayscale image will be described as being composed of 16 rows in the vertical direction × 16 rows in the horizontal direction = 256 pixels.

  First, the CPU 101 converts a color image extracted from the captured image into a grayscale image. For example, the CPU 101 can convert a color image into a grayscale image by employing a weighted average method using NTSC (National Television System Committee) coefficients. For example, if the luminance values of each color in the color image are R, G, and B, and the luminance value in the grayscale image is Y, Y = (0.298912 * R + 0.586611 * G + 0.114478 * B).

  Then, the CPU 101 binarizes the gray scale image. Specifically, the CPU 101 determines whether the luminance value of each pixel constituting the grayscale image is equal to or higher than a threshold value, and binarizes each pixel according to the determination result to generate a binarized image. For example, when each pixel constituting the grayscale image is expressed by 256 gradations from 0 to 255, each pixel constituting the binarized image is expressed by two gradations of 0 or 1. Here, the luminance value of a pixel indicating a dark portion such as the colony 250 is 0 (black), and the luminance value of a pixel indicating a bright portion such as the culture layer 230 is 1 (white).

  For example, the threshold value is set to a luminance value sufficiently higher than the luminance value of the pixel indicating the colony 250 and sufficiently lower than the luminance value of the pixel indicating the culture layer 230. In the present embodiment, processing for selecting a threshold from a threshold range determined based on a grayscale image, and processing for extracting a connected component from a binarized image generated by binarization using the selected threshold, The threshold is repeated from the lowest value (lower limit value) to the highest value (upper limit value) of the threshold range. Here, a process of extracting a connected component from a binarized image generated by binarization using a specific threshold will be described. Note that the threshold range may be determined in advance according to the imaging environment, the type of microorganisms constituting the colony 250, the material of the culture medium, and the like.

  FIG. 4 shows a binarized image. In FIG. 4, pixels indicated by hatching are pixels having a luminance value of 0 (black), and pixels indicated by white portions are pixels having a luminance value of 1 (white). Here, a pixel having a luminance value of 0 (black) is a pixel indicating the colony 250. Accordingly, the pixels having a luminance value of 0 (black) constitute several groups on the binarized image. Each of these groups represents one colony 250 and is hereinafter referred to as a connected component as appropriate. In the example shown in FIG. 4, an example is shown in which three connected components (three colonies 250) indicated by C01, C02, and Co3 are detected.

  Here, it is desirable to manage each connected component by labeling. That is, it is desirable that the connected component can be specified by this label. Hereinafter, the labeling process will be described with reference to FIG. In the present embodiment, an example is described in which 8 connections are used in which the portions that are continuous in the vertical, horizontal, and diagonal directions are the same label. Can also be adopted.

  First, for labeling, the CPU 101 prepares variables in the RAM 103 or the like for storing label numbers corresponding to the number of pixels each corresponding to one of the pixels. Then, the CPU 101 initializes all label numbers to 0. Here, a label number of a certain pixel being 0 indicates that the certain pixel is not labeled, and a label number of a certain pixel being an integer n other than 0 indicates that this certain pixel is an integer n Indicates that it is labeled. In this embodiment, when the luminance value of a pixel is 0 (black), this pixel is labeled. When the luminance value of a pixel is 1 (white), this pixel is not labeled.

  The CPU 101 executes the update process of the label numbers of all the pixels by executing the update process of the label numbers of the target pixels while scanning the target pixels. Note that the target pixel is initially the upper left pixel (the pixel in the first row and the first column) in FIG. 5, and is shifted one by one in the right direction (column direction, horizontal axis direction) to reach the right end. Shift to the pixel at the left end of the next lower row, and similarly shift to the lower right pixel (pixel at the 16th row and the 16th column) in FIG.

  Here, the label number update process will be described. First, when the luminance value of the target pixel is 1 (white), the label number of the target pixel is not updated. On the other hand, when the luminance value of the target pixel is 0 (black), the label number of the target pixel is updated. What the label number of the target pixel is updated depends on what the label numbers of the four pixels of the upper left pixel, the upper pixel, the upper right pixel, and the left pixel are based on the target pixel. For example, when the pixel in the third row and the fourth column (hereinafter referred to as “P34”) indicated by P34 in FIG. 5 is the target pixel, the label of P34 is set according to the label numbers of P23, P24, P25, and P33. The number is determined.

  Specifically, when the label numbers of all four pixels are 0, a new label number is assigned to the target pixel. That is, a label number that is one larger than the largest label number among the label numbers assigned in the past is set as the label number of the target pixel. On the other hand, if any of the label numbers of the four pixels is not 0, the smallest label number among the four label numbers (however, a label number other than 0) is set as the label number of the target pixel. In addition, when there is a label number other than the smallest label number among the four label numbers (however, a label number other than 0), all of the pixels to which a label number other than the smallest label number is assigned. Rewrite the label number to this smallest label number.

  When the label number updating process described above is executed for all pixels, the labeling process is completed. FIG. 5 shows an example in which three connected components C01, C02, and C03 are detected by the labeling process. Since label numbers rewritten with other label numbers are not used, it is desirable to reassign the label numbers so that the label numbers used are serial numbers.

  As shown in FIG. 5, the connected component is indefinite. Therefore, in this embodiment, the connected component is treated as an indefinite type until at least the connected component counting process described later is completed. However, when the connected component is approximated by a rectangle, the process of determining the size of the connected component and the like are facilitated. Therefore, for example, after the connected component counting process is completed, the connected component can be approximated by a rectangle circumscribing the connected component as shown in FIG. FIG. 6 shows an example in which the connected component indicated by C01 is approximated by C11, the connected component indicated by C02 is approximated by C12, and the connected component indicated by C03 is approximated by C13. Note that the rectangle is surrounded by two line segments parallel to the vertical axis direction (row direction) in the captured image and two line segments parallel to the horizontal axis direction (column direction) in the captured image. Note that Ly indicates the length of the rectangle indicated by C11 in the vertical axis direction, and Lx indicates the length of the rectangle indicated by C11 in the horizontal axis direction. In this example, the length of the short side of the rectangle indicated by C11 is Lx, and the length of the long side of the rectangle indicated by C11 is Ly.

  Next, a functional configuration of the colony counting apparatus 100 according to the present embodiment will be described with reference to FIG. As shown in FIG. 7, the colony counting apparatus 100 includes a grayscale image acquisition unit 11, a binarized image generation unit 12, a connected component extraction unit 13, a connected component information addition unit 14, a threshold supply unit 15, and a colony counting unit 16. , A threshold range determination unit 17, a color image acquisition unit 18, a hue range specification unit 19, and a gray scale image generation unit 20.

  The gray scale image acquisition unit 11 acquires a gray scale image showing the state of the culture medium after culturing the microorganisms contained in the specimen. The gray scale image acquisition unit 11 includes, for example, a CPU 101.

  The binarized image generating unit 12 generates a binarized image by binarizing the grayscale image acquired by the grayscale image acquiring unit 11 with the threshold every time the threshold is supplied. The binarized image generation unit 12 includes, for example, a CPU 101.

  The connected component extraction unit 13 extracts a connected component from the binarized image by performing a labeling process on the binarized image every time the binarized image generating unit 12 generates the binarized image. The connected component extraction unit 13 includes, for example, a CPU 101.

  The connected component information adding unit 14 represents a connected component that does not include the connected component indicated by the connected component information stored in the storage unit 30 among the connected components every time the connected component is extracted by the connected component extracting unit 13. The component information is further stored in the storage unit 30. The connected component information adding unit 14 includes, for example, a CPU 101.

  The threshold supply unit 15 increases the threshold value by a predetermined value and supplies it to the binarized image generation unit 12 every time the processing of the connected component information adding unit 14 is completed until the threshold value reaches the upper limit value. To do. The threshold supply unit 15 includes, for example, a CPU 101.

  The colony counting unit 16 counts the number of colonies of microorganisms based on the connected component information stored in the storage unit 30. The colony counting unit 16 includes, for example, a CPU 101.

  The threshold range determination unit 17 determines the lower limit value and the upper limit value based on the distribution of the luminance values of the pixels constituting the grayscale image.

  For example, the threshold range determination unit 17 determines the luminance value having the highest frequency in the luminance range of the maximum mountain-shaped frequency distribution included in the luminance histogram representing the frequency for each luminance value of the pixels constituting the grayscale image. Identify. Then, the threshold range determination unit 17 determines the highest luminance value as the upper limit value among the luminance values that are lower than the specified luminance value and are equal to or less than a predetermined ratio with respect to the frequency of the specified luminance value. To do.

  For example, the threshold range determination unit 17 specifies the minimum luminance value of the luminance values of the pixels constituting the grayscale image, and determines the specified luminance value as the lower limit value. The colony counting unit 16 includes, for example, a CPU 101.

  The color image acquisition unit 18 acquires a color image obtained by imaging the culture medium. For example, the color image is extracted from the captured image. The captured image may be a still image or a moving image. The color image acquisition unit 18 includes, for example, a CPU 101.

The hue range specifying unit 19 specifies the hue range of the colony based on the hue distribution of the pixels constituting the color image acquired by the color image acquiring unit 18.
For example, the hue range specifying unit 19 removes the hue range of the maximum mountain-shaped frequency distribution from the hue histogram representing the frequency for each hue of the pixels constituting the color image, and thus the hue range of the mountain-shaped frequency distribution exists. As the colony hue range. The hue range specifying unit 19 includes, for example, a CPU 101.

  The gray scale image generation unit 20 generates a gray scale image to be acquired by the gray scale image acquisition unit 11. This grayscale image is, for example, the luminance of pixels that are not included in the hue range specified by the hue range specifying unit 19 among the pixels constituting the image obtained by converting the color image acquired by the color image acquiring unit 18 to grayscale. It is obtained by setting the value to a luminance value higher than the upper limit value.

  Further, the gray scale image generation unit 20 causes the gray scale image acquisition unit 11 to acquire a gray scale image obtained by applying a Gaussian filter to an image obtained by converting the color image acquired by the color image acquisition unit 18 into a gray scale. It may be generated as a gray scale image. The gray scale image generation unit 20 includes a CPU 101, for example.

  The storage unit 30 stores connected component information indicating the connected component extracted by the connected component extracting unit 13. The connected component information stored in the storage unit 30 is updated by the CPU 101 or the like so that each time the connected component is extracted by the connected component extracting unit 13, the extracted connected component is further included. The connected component information is, for example, information indicating the position (coordinates) of each pixel constituting the connected component and the label number of the connected component, and is information managed by an array variable or the like. The storage unit 30 includes, for example, a RAM 103 and a storage device 104.

  Next, a colony counting process executed by the colony counting apparatus 100 according to the present embodiment will be described. FIG. 8 is a flowchart showing a colony counting process executed by the colony counting apparatus 100. The colony counting device 100 starts the colony counting process shown in FIG. 8 when detecting that the input device 105 has received an instruction to start the colony counting process by the operator, for example.

  First, the CPU 101 acquires a captured image (step S101). The acquisition source of the captured image is typically the imaging device 108, but may be an external device or the like with which the storage device 104, the image storage unit 109, and the communication device 106 can communicate. The captured image is typically a color image obtained by capturing the dedicated sheet 200. The captured image may be a monochrome image. Note that the CPU 101 may execute various trimming processes and filtering processes on the acquired captured image. For example, the CPU 101 stores the acquired captured image in the RAM 103.

  When completing the process in step S101, the CPU 101 acquires a color image (step S102). For example, the CPU 101 extracts a rectangular area from the captured image and generates a color image. The CPU 101 stores the acquired color image in the RAM 103.

  When completing the process in step S102, the CPU 101 generates a grayscale image by converting the acquired color image into a grayscale (step S103). For the gray scale conversion, for example, a weighted average method using NTSC counting can be adopted. The generated grayscale image is stored in the RAM 103, for example.

  When completing the process of step S103, the CPU 101 executes a Gaussian filtering process (step S104). Note that the Gaussian filter is a filter that increases the weight when calculating the average value for pixels closer to the target pixel and decreases the weight for calculating the average value for pixels farther from the target pixel. This weight can be calculated using a Gaussian distribution function.

  FIG. 9 shows a 5 × 5 operator (kernel) used for the Gaussian filter. The grayscale image has a smooth luminance value distribution by Gaussian filtering. FIG. 10A shows a grayscale image when the Gaussian filtering process is not executed. FIG. 10A shows an example in which one colony 250 is detected as three connected components indicated by C21, C22, and C23 because the luminance value distribution of pixels representing the colony 250 is not uniform. Show.

  FIG. 10B shows a grayscale image when the Gaussian filtering process is executed. FIG. 10B shows an example in which one colony 250 is detected as one connected component indicated by C20 because the variation in the luminance value of the pixel representing the colony 250 is reduced by the Gaussian filtering process. ing. When the CPU 101 completes the Gaussian filtering process in step S104, the CPU 101 executes a threshold range determination process (step S105). The threshold range determination process will be described in detail with reference to FIG. FIG. 11 is a flowchart showing a threshold range determination process executed by the colony counting device 100. The threshold range is determined by the lower limit value of the threshold range and the upper limit value of the threshold range.

  First, the CPU 101 generates a histogram of luminance values of a grayscale image (step S201). For example, assuming that the gradation of the pixels constituting the grayscale image is 256 gradations of 0 to 255, the CPU 101, for each of the luminance values of 0 to 255, among the pixels constituting the grayscale image, The number of pixels (frequency) that is the luminance value is obtained.

  When the CPU 101 completes the process of step S201, the CPU 101 applies a moving average to the generated histogram (step S202). By such moving average processing, noise components on the histogram are removed, and a smooth histogram is generated. Information representing the generated histogram is stored in the RAM 103 as appropriate.

  FIG. 12 shows a histogram of luminance values of a grayscale image. The gray scale image is an image representing the culture layer 230 (medium) in which the colony 250 is generated and the base sheet 210 (mounting paper). Therefore, as shown in FIG. 12, the frequency of the pixel of the luminance value representing the culture layer 230 and the frequency of the pixel of the luminance value representing the base sheet 210 are increased. That is, the luminance value histogram of the gray scale image has a large peak in the luminance value of the pixel representing the culture medium and the luminance value of the pixel representing the mount. Note that when the mount is not included in the color image, no peak occurs in the luminance value of the pixel representing the mount. On the other hand, when a ruled line is included in the color image, a peak occurs in the luminance value of the pixel representing the ruled line.

  Here, the colony 250 has a darker color than the culture layer 230 and the substrate sheet 210. Therefore, the luminance value of the pixel representing the colony 250 is lower than the luminance value of the pixel representing the culture layer 230 and the luminance value of the pixel representing the base sheet 210. In addition, since the color of the colony 250 is not uniform, the luminance value of the pixel representing the colony 250 has a certain width. Similarly, the luminance value of the pixel representing the culture layer 230 and the luminance value of the pixel representing the culture layer 230 also have a certain range.

  In the present embodiment, the lower limit value indicating the minimum value of the threshold range for binarization is set as the minimum value of the luminance value of the pixel representing the colony 250. That is, the frequencies are checked in order of increasing brightness value from 0, and the brightness value at which the frequency is not 0 for the first time is set as the lower limit value.

  First, when the process of step S202 is completed, the CPU 101 sets the target luminance value to 0 (step S203). Specifically, the CPU 101 provides a variable representing the noticed luminance value in the RAM 103 and initializes the value of the variable representing the noticed luminance value to 0.

  When the process of step S203 is completed, the CPU 101 determines whether or not the frequency of the target luminance value is 0 (step S204). Note that the CPU 101 can confirm the frequency of the target luminance value by referring to the data representing the histogram described above.

  When the CPU 101 determines that the frequency of the target luminance value is 0 (step S204: YES), the CPU 101 increments the target luminance value (step S205). For example, the CPU 101 writes a value that is one greater than the current value in a variable representing the target luminance value stored in the RAM 103. Thus, the increment amount is typically 1, but is not limited to 1. When the CPU 101 completes the process of step S205, the process returns to step S204.

  On the other hand, if the CPU 101 determines that the frequency of the target luminance value is not 0 (step S204: NO), the CPU 101 sets the target luminance value to the lower limit value (step S206). Specifically, the CPU 101 provides a variable indicating the lower limit value on the RAM 103, and substitutes the variable value indicating the luminance value of interest into this variable.

  In the present embodiment, the upper limit value indicating the maximum value of the threshold range for binarization is determined from the luminance value having the highest frequency among the luminance values of the pixels representing the culture medium (hereinafter referred to as “peak luminance value”). Is the lowest luminance value, and the maximum luminance value is equal to or less than a predetermined ratio of the frequency of the peak luminance value. That is, the frequencies are checked in order of descending power from the peak luminance value, and the luminance value that becomes the first processing frequency or less of the frequency of the peak luminance value is set as the upper limit value.

  First, when the CPU 101 completes the process of step S206, the CPU 101 specifies the maximum mountain area in the histogram (step S207). This maximum mountain region is configured by the luminance value of the pixel representing the culture medium. This is because the number of pixels in the luminance range representing the culture medium is the largest on the grayscale image. If the number of pixels representing the medium (medium area) is larger than the number of pixels representing the mount (mounting area) on the grayscale image, a mountain region constituted by the luminance values of the pixels representing the mount Rather, the mountain-shaped region constituted by the luminance values of the pixels representing the culture medium becomes larger. A method for detecting the maximum mountain region is arbitrary.

  For example, while scanning the target luminance value in order of increasing from 0, the frequency of the target luminance value and the total frequency of a predetermined number of luminance values in the vicinity of the target luminance value are compared, and the target luminance value having the maximum total frequency is compared. Temporarily, the luminance value at the center of the maximum mountain region can be determined. It should be noted that a range of luminance values in which the ratio of the central luminance value with respect to the total frequency is a predetermined ratio or more can be determined as a mountain region.

  On the histogram, the frequency peak of the pixel representing the mount may be higher than the frequency peak of the pixel representing the culture medium. However, on the grayscale image, when the number of pixels in the luminance range representing the culture medium is the largest, by applying a strong moving average to the histogram, the frequency peak of the pixel representing the mount is made higher than the frequency peak of the pixel representing the culture medium. Can be lowered. Therefore, by appropriately applying a moving average to the histogram, it is possible to easily find a mountain-shaped region constituted by the peak of the frequency of the pixel representing the culture medium or the luminance value of the pixel representing the culture medium.

  When the CPU 101 completes the process of step S207, it specifies the luminance value of the maximum frequency in the maximum mountain area (step S208). That is, the CPU 101 specifies the luminance value of the maximum frequency from the luminance range that forms the maximum mountain area. A variable indicating the luminance value of the maximum frequency is stored in the RAM 103 or the like. In FIG. 12, the maximum frequency in the maximum mountain region is indicated by Nmax. The frequency at the upper limit is a frequency of Nmax × 20%.

  When completing the process of step S208, the CPU 101 sets the luminance value of interest to the luminance value of the maximum frequency (step S209). That is, the CPU 101 initializes the value of the variable indicating the target luminance value stored in the RAM 103 to the luminance value of the maximum frequency.

  When completing the process in step S209, the CPU 101 determines whether or not the frequency of the target luminance value is equal to or less than a predetermined ratio of the maximum frequency (step S210). The predetermined ratio can be, for example, 20%, but is not limited to 20%.

  If the CPU 101 determines that the frequency of the target luminance value is not equal to or less than the predetermined ratio of the maximum frequency (step S210: NO), the CPU 101 decrements the target luminance value (step S211). That is, the CPU 101 substitutes a value smaller than the current value by one for the variable indicating the target luminance value stored in the RAM 103. The value to be decremented is typically 1, but is not limited to 1. When completing the process in step S211, the CPU 101 returns the process to step S210.

  On the other hand, when the CPU 101 determines that the frequency of the target luminance value is equal to or less than the predetermined ratio of the maximum frequency (step S210: YES), the CPU 101 sets the target luminance value to the upper limit value (step S212). Specifically, the CPU 101 provides a variable indicating the upper limit value on the RAM 103, and substitutes the variable value indicating the target luminance value into this variable. When completing the process in step S212, the CPU 101 completes the threshold range determination process.

  CPU101 will perform a hue band pass filtering process, if the threshold range determination process of step S105 is completed (step S106). The hue bandpass filtering process will be described in detail with reference to FIG. FIG. 13 is a flowchart showing the hue bandpass filtering process executed by the colony counting apparatus 100.

  First, the CPU 101 generates a hue histogram of a color image (step S301). Specifically, the CPU 101 determines the hue of each pixel constituting the color image based on the ratio of the R (red) luminance value, the G (green) luminance value, and the B (blue) luminance value. decide. The hue is represented by red → orange → yellow → green → blue → indigo → purple → red, and 0 to 360 degrees. FIG. 14 shows a hue histogram of a color image.

  As shown in FIG. 14, on the histogram, a mountain-shaped area constituted by the hue of the pixel representing the culture medium and a mountain-shaped area constituted by the hue of the pixel representing the colony 250 are seen. In the present embodiment, the mountain-shaped region constituted by the hue of the pixel representing the culture medium is excluded from the histogram, and the mountain-shaped region constituted by the hue of the pixel representing the colony 250, which is seen on the histogram after the exclusion, Set to a valid hue range. That is, pixels in a hue range other than the effective hue range are invalidated.

  Such hue bandpass filtering processing makes it difficult for the texture and ruled lines of the culture medium to be erroneously recognized as colonies 250.

  First, the CPU 101 generates a hue histogram of a color image (step S301). Specifically, the CPU 101 determines the hue for each pixel constituting the color image based on the ratio of the luminance values of the three primary colors. Then, the CPU 101 totals the number of pixels of this hue for each hue and obtains the frequency of this hue.

  When the CPU 101 completes the process of step S301, the CPU 101 identifies the maximum mountain area in the histogram (step S302). As in step S207, the CPU 101 can obtain the maximum mountain area in the histogram.

  When the CPU 101 completes the process of step S302, the CPU 101 identifies the mountain area after the removal of the largest mountain area as the hue range of the colony 250 (step S303). The CPU 101 can obtain a mountain region in the histogram after the removal of the largest mountain region by the same process as in step S302. The CPU 101 stores information indicating the specified hue range in the RAM 103.

  When the CPU 101 completes the process of step S303, the CPU 101 sets the maximum luminance value to the pixels representing the hue outside the hue range of the colony 250 among the pixels constituting the grayscale image (step S304). Here, the maximum value of the luminance value is, for example, 255 (white) when the grayscale image has 256 gradations. According to such processing, regardless of the threshold value for binarizing the grayscale image, the pixel for which the maximum luminance value has been set is set to 1 (white) in the binarized image. Is set. When completing the process in step S304, the CPU 101 completes the hue bandpass filtering process.

  When the CPU 101 completes the hue bandpass filtering process in step S106, the CPU 101 executes a connected component count process (step S107). The connected component counting process will be described in detail with reference to FIG. FIG. 15 is a flowchart showing a connected component counting process executed by the colony counting apparatus 100.

  First, the CPU 101 initializes connected component information (step S401). The connected component information is stored in the storage device 104 and managed as an array variable. The connected component information is indicated by, for example, the position and size of the connected component, the label number assigned to the connected component, and the like for each detected connected component. When the connected component information does not exist in the storage device 104, the CPU 101 does not have to initialize the connected component information.

  In the present embodiment, the connected component is treated as being indefinite as shown as C01, C02, and C03 in FIG. For example, one connected component can be represented by a structure type variable (hereinafter referred to as “structure variable”) including an array of coordinates of each pixel constituting the connected component and a label number. The connected component information is managed by, for example, an array of structure variables corresponding to the number of detected connected components.

  CPU101 will set a lower limit to a threshold value, if the process of step S401 is completed (step S402). This threshold is a threshold used when converting a grayscale image into a binarized image. The lower limit value is a lower limit value of the threshold value, and is set in step S206. For example, the CPU 101 substitutes the value of the variable indicating the lower limit value provided in the RAM 103 into the variable indicating the threshold value provided in the RAM 103.

  When completing the process in step S402, the CPU 101 binarizes the grayscale image with the set threshold (step S403). The CPU 101 stores information representing the generated binarized image in the RAM 103.

  When completing the process in step S403, the CPU 101 executes a labeling process on the binarized image (step S404). The labeling process is the process described with reference to FIGS. Note that zero or more connected components are detected by the labeling process. The detected connected component is given a unique label number. The connected component is managed by the structure variable described above. Information indicating the connected component is appropriately stored in the RAM 103 or the like.

  When completing the process in step S404, the CPU 101 determines whether there is an unselected connected component (step S405). Specifically, the CPU 101 determines whether there is a connected component that has not been selected in step S406 among the connected components detected by the labeling process in step S404, with reference to a flag associated with the connected component, or the like. Can be determined. When connected components are selected in order from the smallest label number in step S406, it is determined whether there is an unselected connected component depending on whether the connected component with the largest label number is selected in step S406. can do.

  When determining that there is an unselected connected component (step S405: YES), the CPU 101 selects one connected component (step S406). Note that the CPU 101 can select the connected components in the order of increasing label numbers or in descending order of label numbers. In this case, for example, the CPU 101 stores a variable indicating the label number of the last selected connected component in the RAM 103.

  When completing the process in step S406, the CPU 101 determines whether or not the selected connected component includes the detected connected component (step S407). Specifically, the CPU 101 selects each of all the pixels constituting the connected component from among the connected components detected by the labeling process of the binarized image generated by the binarization using the threshold value lower than the current value. It is determined whether or not there is a connected component that matches any of all the pixels constituting the connected component. In addition, the connected component information which shows the connected component detected in the past is memorize | stored in the memory | storage device 104, for example.

  If the CPU 101 determines that the detected connected component is included (step S407: YES), the CPU 101 returns the process to step S405. On the other hand, when the CPU 101 determines that the detected connected component is not included (step S407: NO), the CPU 101 adds the connected component to the labeling result (step S408). Specifically, the CPU 101 adds information indicating the selected connected component to the end of the array of connected component information stored in the storage device 104. Alternatively, the CPU 101 updates the connected component information so that the connected component information further includes information indicating the selected connected component.

  When the CPU 101 completes the process of step S408, it reassigns the label number (step S409). That is, the CPU 101 gives unique label numbers to all the connected components that have already been detected and one connected component that is newly added this time. The method for assigning the label number can be adjusted as appropriate. For example, the CPU 101 can assign label numbers in the ascending order (descending order) in the order in which the connected components are arranged or detected. When the CPU 101 completes the process of step S409, the process returns to step S405.

  On the other hand, when determining that there is no unselected connected component (step S405: NO), the CPU 101 increments the threshold (step S410). Note that it is determined that there is no unselected connected component that all the connected components to be added among the connected components detected from the binarized image binarized by the current threshold are added. Means.

  The increment amount of the threshold can be adjusted as appropriate. For example, the increment amount can be 1 or other predetermined integer value. Alternatively, for example, the increment amount may be a quotient obtained by dividing the difference between the lower limit value of the threshold and the upper limit value of the threshold by a predetermined integer. The CPU 101 substitutes a value indicating the incremented threshold value into a variable indicating the threshold value stored in the RAM 103.

  When completing the process in step S410, the CPU 101 determines whether or not the threshold value exceeds the upper limit value (step S411). That is, the CPU 101 compares the threshold value stored in the RAM 103 with the variable indicating the upper limit value to determine whether or not the incremented threshold value in step S410 exceeds the upper limit value set in step S212. To determine.

  If the CPU 101 determines that the threshold value does not exceed the upper limit value (step S411: NO), it returns the process to step S403. On the other hand, if the CPU 101 determines that the threshold value exceeds the upper limit value (step S411: YES), it saves the labeling result (step S412). The CPU 101 can obtain a labeling result based on the connected component information stored in the storage device 104, and can store information indicating the labeling result in the storage device 104 or the like. When completing the process in step S412, the CPU 101 completes the connected component count process.

  Here, with reference to FIGS. 16 to 18, a state in which connected components are added by the connected component counting process will be described.

  FIG. 16A shows a grayscale image. This grayscale image is captured by three colonies 250, a colony 250 surrounded by an indefinite shape indicated by C31, a colony 250 surrounded by a rectangle indicated by C32, and a colony 250 surrounded by a rectangle indicated by C33. It has been shown. Here, the colony 250 surrounded by the irregular shape indicated by C31 has a dark color, and the colony 250 surrounded by the rectangle indicated by C32 has a normal dark color, and is surrounded by the rectangle indicated by C33. Assume that the colony 250 has a light color.

  FIG. 16B is a diagram illustrating connected components detected when a binarized image obtained by binarizing a grayscale image with the threshold set to the lowest value is labeled. Here, in the binarization when the threshold value is set to the lowest value, only pixels having luminance values indicating dark colors are set to 0 (black), and pixels having luminance values indicating other dark colors are set to 1 (white). ). In this case, as shown in FIG. 16B, only the colony 250 having a dark color surrounded by the rectangle of C41 is detected as a connected component.

  FIG. 17A shows connected components detected when a binarized image obtained by binarizing a grayscale image by setting the threshold to a threshold increased by one unit from the lowest value is labeled. FIG. Here, in the binarization when the threshold value is set to a threshold value that is increased by one unit from the lowest value, the pixel of the luminance value indicating the dark color and the pixel of the luminance value indicating the normal dark color are 0 (black). ) And a pixel having a luminance value indicating a light color is 1 (white). In this case, as shown in FIG. 17A, a colony 250 having a dark color surrounded by an indefinite shape of C42 and a colony 250 having a normal dark color surrounded by a rectangle of C43 are connected. Detected as a component. However, since the connected component indicated by the C42 indefinite shape includes the connected component indicated by the already detected C41 rectangle, it is not added to the connected component information as a newly detected connected component. On the other hand, since the connected component indicated by the rectangle of C43 does not include any already detected connected component, it is added to the connected component information as a newly detected connected component.

  FIG. 17B is a diagram illustrating connected components detected when a binarized image obtained by binarizing a grayscale image by setting a threshold value after an increment of 1 unit is labeled. Here, in the binarization when the threshold value after incrementing by one unit is set, the luminance value pixel indicating the dark color and the luminance value pixel indicating the normal dark color are set to 0 (black) and light. A pixel having a luminance value indicating a color is 1 (white). In this case, as shown in FIG. 17B, a colony 250 having a dark color surrounded by an indefinite shape of C51 and a colony 250 having a normal dark color surrounded by a rectangle of C52 are connected. Detected as a component.

  FIG. 18A is a diagram showing connected components detected when a binarized image obtained by binarizing a grayscale image by setting the threshold to a threshold value further increased by 1 unit is labeled. is there. Here, in the binarization when the threshold value is set to a threshold value further increased by one unit, a luminance value pixel indicating a dark color, a luminance value pixel indicating a normal color, and a light value indicating a light color. It is assumed that these pixels are 0 (black). In this case, as shown in FIG. 18A, a combination of a colony 250 having a dark color and a colony 250 having a normal dark color surrounded by an indefinite shape of C44 and a rectangle of C45 The enclosed colony 250 having a normal dark color is detected as a connected component. However, the connected component indicated by the indefinite shape of C44 includes the connected component indicated by the rectangle of C41 that has already been detected and the connected component indicated by the rectangle of C43. Not added to connected component information. On the other hand, since the connected component indicated by the rectangle of C45 does not include any already detected connected component, it is added to the connected component information as a newly detected connected component.

  FIG. 18B is a diagram illustrating connected components detected when a binarized image obtained by binarizing a grayscale image by setting a threshold after an increase of 2 units is labeled. Here, in the binarization when the threshold value after an increase of 2 units is set, a pixel with a luminance value indicating a dark color, a pixel with a luminance value indicating a normal color, and a pixel with a luminance value indicating a light color, Is set to 0 (black). In this case, as shown in FIG. 18B, a combination of a colony 250 having a dark color and a colony 250 having a normal dark color surrounded by an indefinite shape of C61 and a rectangle of C62 A surrounded colony 250 having a light color is detected as a connected component.

  As shown in FIG. 16B, FIG. 17A, and FIG. 18A, when the connected component is accumulated while increasing the threshold value by one unit by the connected component counting process in this embodiment, Three connected components, a connected component indicated by a rectangle, a connected component indicated by a rectangle C43, and a connected component indicated by a rectangle C45, are appropriately detected. That is, in this case, the three colonies 250 having different color densities can be detected appropriately.

  On the other hand, as shown in FIGS. 16 (B), 17 (B), and 18 (B), if a threshold is set only once and a connected component is extracted, no matter how the threshold is set. The three colonies 250 having different color densities cannot be detected properly.

  When the CPU 101 completes the connected component counting process in step S107, the CPU 101 executes a count result display process (step S108). Specifically, based on the connected component information stored in the storage device 104, the CPU 101 displays the position, size, shape, and number of connected components on the screen together with the color image. To do.

  FIG. 19 is a diagram showing a screen presented by the count result display process. On the screen 300, a number image 320 indicating the number of detected connected components is superimposed on a background image 310 indicating a part of the color image, and a detected position indicating the position, size, shape, etc. of the detected connected components. A mark 330 is displayed. FIG. 19 shows an example in which the detected irregular connected component is approximated to a rectangular connected component and displayed as a detection position mark 330. FIG. 19 shows an example in which a part of the color image is enlarged and displayed as the background image 310 on the screen 300, but the entire color image may be displayed as the background image 310 on the screen 300. . As described above, the screen 300 can easily present the position and shape of the colony 250 on the color image corresponding to the detected connected component to the operator. The operator can appropriately observe the presented connected component by appropriately operating the input device 105 or the like to enlarge, reduce, or scroll the color image.

  Note that the colony counting apparatus 100 may not only display information on the detected connected components on the screen 300 but also store the information in the storage device 104 or the like, or may be connected to the network via the communication device 106. You may transmit to an apparatus etc.

  When completing the process of step S108, the CPU 101 completes the colony counting process.

  According to the present embodiment, it is possible to appropriately count the number of colonies 250 having variations in color density. In addition, on the gray scale image, it is expected that the color around the dark colony 250 becomes lighter as the distance from the colony 250 increases. Therefore, when the binarized luminance value is gradually increased, the connected component already detected by the detection process using the low threshold is detected again by the detection process using the high threshold. However, according to the present embodiment, when the connected component newly detected by the detection process with the high threshold includes the connected component already detected by the detection process with the low threshold, the newly detected connected component is It is not considered newly detected. For this reason, even if the labeling process is executed with a plurality of threshold values, the connected components can be made difficult to be counted repeatedly. Therefore, the number of colonies 250 indicated by the connected component can be accurately counted.

(Modification)
The present invention is not limited to the one disclosed in the above embodiment.

  For example, the method for determining the lower limit value of the threshold and the upper limit value of the threshold is not limited to the method disclosed in the above embodiment. For example, the lower limit value of the threshold value and the upper limit value of the threshold value can be appropriately determined by various methods based on the gray scale image. For example, when the range of the luminance value of the pixel representing the colony 250 can be predicted in advance, the threshold range may be determined in advance.

  Further, the method of the hue bandpass filtering process is not limited to the method disclosed in the above embodiment. For example, the hue band pass filtering process can be performed on the grayscale image by various methods based on the color image. Further, the hue band pass filtering process may not be performed.

  Further, the Gaussian filtering processing technique is not limited to the technique disclosed in the above embodiment. A Gaussian filtering process can be applied to the grayscale image by various methods. Further, there is no need for Gaussian filtering processing. Further, a moving average filter process or the like may be executed instead of the Gaussian filtering process. Further, the timing at which the Gaussian filtering process is executed is not limited to the example of the above embodiment. For example, the Gaussian filtering process may be executed after the hue bandpass filtering process.

  Moreover, the culture medium applied to this invention is not limited to what was disclosed in the said embodiment. For example, a film medium of a type without a mount can be employed. FIG. 20 shows a state in which the entire surface of the base sheet 211 is a culture medium (culture layer), and microbial colonies 250 are generated on the entire surface of the base sheet 211. In addition, the base sheet 211 has identification information 220 regarding the culture medium. Furthermore, ruled lines 260 are drawn at regular intervals in the vertical direction and the horizontal direction on the base material sheet 211 (medium).

  In this case, for example, a color image is acquired by imaging a region 240 that is a partial region of the base sheet 211. In the color image, a base sheet 211 (medium), a colony 250, a ruled line 260, and the like are expressed. In this case, the luminance value of each color of RGB is different from the color of the base sheet 211 (medium), the color of the colony 250, and the color of the ruled line 260. Therefore, the colony 250 can be distinguished from the base material sheet 211 (medium) and the ruled line 260 by hue filtering processing and binarization processing.

  In the first embodiment, the colony counting device 100 includes the CPU 101, the ROM 102, and the RAM 103, and the CPU 101 shows an example in which the colony counting process is realized by software in accordance with the program stored in the ROM 102. However, the colony counting process executed by the colony counting apparatus 100 is not limited to that realized by software. For example, the colony counting apparatus 100 may be configured by a microcomputer, an FPGA (Field Programmable Gate Array), a PLD (Programmable Logic Device), a DSP (Digital Signal Processor), or the like.

  Note that the colony counting apparatus according to the present invention can be realized by using a normal computer system without using a dedicated system. For example, a computer readable recording such as a flexible disk, a CD-ROM (Compact Disk-Read Only Memory), a DVD (Digital Versatile Disk), or a MO (Magnet Optical Disk) is stored in a computer. You may comprise the colony counting device which performs the above-mentioned process by memorize | storing and distributing to a medium and installing this in a computer system. Furthermore, the program may be stored in a disk device or the like included in a server device on the Internet, and may be downloaded onto a computer by being superimposed on a carrier wave, for example.

DESCRIPTION OF SYMBOLS 11 Gray scale image acquisition part 12 Binary image generation part 13 Connected component extraction part 14 Connected component information addition part 15 Threshold supply part 16 Colony counting part 17 Threshold range part 18 Color image acquisition part 19 Hue range identification part 20 Gray scale image Generation unit 30 Storage unit 100 Colony counting device 101 CPU
102 ROM
103 RAM
104 Storage device 105 Input device 106 Communication device 107 Imaging control unit 108 Imaging device 109 Image storage unit 110 Display control unit 111 Display device 112 Audio processing unit 113 Audio output device 120 Bus 200 Dedicated sheet 210, 211 Base sheet 220 Identification information 230 Culture layer 240 Region 250 Colony 260 Ruled line 300 Screen 310 Background image 320 Number image 330 Detection position mark

Claims (9)

A grayscale image acquisition means for acquiring a grayscale image showing the state of the culture medium after culturing the microorganisms contained in the specimen;
Binarized image generating means for generating a binarized image by binarizing the grayscale image acquired by the grayscale image acquiring means by the threshold value every time a threshold is supplied;
Each time the binarized image is generated by the binarized image generating means, a labeling process is performed on the binarized image, thereby connecting a group of pixels indicating one colony from the binarized image. Connected component extraction means for extracting the components;
Each time the connected component is extracted by the connected component extraction means, the storage device stores connected component information indicating a connected component that does not include the connected component indicated by the connected component information stored in the storage device. Connected component information adding means for further storing
Threshold supply means for increasing the threshold value by a predetermined value and supplying it to the binarized image generating means each time processing of the connected component information adding means is completed until the threshold value reaches the upper limit value. When,
Colony counting means for counting the number of colonies of the microorganism based on the connected component information stored in the storage device,
A colony counting device characterized by that. Threshold range determining means for determining the lower limit value and the upper limit value based on a distribution of luminance values of pixels constituting the gray scale image,
The colony counting device according to claim 1, wherein: The threshold range determination means identifies a luminance value having the highest frequency in the luminance range of the maximum mountain-shaped frequency distribution included in the luminance histogram representing the frequency for each luminance value of the pixels constituting the grayscale image. The highest luminance value is determined as the upper limit value among luminance values that are lower than the specified luminance value and that are less than a predetermined ratio with respect to the frequency of the specified luminance value.
The colony counting device according to claim 2, wherein The threshold range determining means specifies the minimum luminance value of the luminance values of the pixels constituting the grayscale image, and determines the specified luminance value as the lower limit value;
The colony counting device according to claim 2 or 3, wherein Color image acquisition means for acquiring a color image obtained by imaging the culture medium;
A hue range specifying means for specifying a hue range of the colony based on a hue distribution of pixels constituting the color image acquired by the color image acquiring means;
Among the pixels constituting the image obtained by converting the color image acquired by the color image acquisition unit into a gray scale, the luminance value of the pixel not included in the hue range specified by the hue range specifying unit is determined from the upper limit value. Grayscale image generation means for generating a grayscale image acquired by setting a higher brightness value as a grayscale image to be acquired by the grayscale image acquisition means,
The colony counting apparatus according to any one of claims 1 to 4, wherein the colony counting apparatus is characterized in that The hue range specifying means obtains a hue range of a mountain-shaped frequency distribution that exists after excluding the hue range of the maximum mountain-shaped frequency distribution from a hue histogram representing the frequency for each hue of pixels constituting the color image. Specify as the hue range of the colony,
The colony counting device according to claim 5, wherein Color image acquisition means for acquiring a color image obtained by imaging the culture medium;
Grayscale image generation for generating a grayscale image obtained by applying a Gaussian filter to a grayscale image obtained by converting the color image acquired by the color image acquisition unit to the grayscale image acquisition unit Means further comprising:
The colony counting device according to any one of claims 1 to 6, wherein A grayscale image acquisition step of acquiring a grayscale image showing the state of the culture medium after culturing the microorganisms contained in the specimen;
A binarized image generating step of generating a binarized image by binarizing the grayscale image acquired by the grayscale image acquiring step by the threshold value every time a threshold is supplied;
Each time the binarized image is generated by the binarized image generating step, a labeling process is performed on the binarized image, thereby connecting a group of pixels indicating one colony from the binarized image. A connected component extraction step for extracting components;
Each time the connected component is extracted by the connected component extracting step, connected storage information indicating connected components that do not include the connected components indicated by the connected component information stored in the storage device among the connected components is stored in the storage device. Connected component information adding step to be further stored in
A threshold supply step of increasing the threshold value by a predetermined value and executing the binarized image generation step each time the processing of the connected component information adding step is completed until the threshold value reaches the upper limit value. When,
A colony counting step for counting the number of colonies of the microorganism based on the connected component information stored in the storage device,
The colony counting method characterized by this. Computer
A grayscale image acquisition means for acquiring a grayscale image showing the state of the culture medium after culturing the microorganisms contained in the specimen;
A binarized image generating unit that generates a binarized image by binarizing the grayscale image acquired by the grayscale image acquiring unit by the threshold every time a threshold is supplied;
Each time the binarized image is generated by the binarized image generating means, a labeling process is performed on the binarized image, thereby connecting a group of pixels indicating one colony from the binarized image. Connected component extraction means for extracting components;
Each time the connected component is extracted by the connected component extraction means, the storage device stores connected component information indicating a connected component that does not include the connected component indicated by the connected component information stored in the storage device. Connected component information adding means for further storing
Threshold supply means for increasing the threshold value by a predetermined value and supplying it to the binarized image generating means each time processing of the connected component information adding means is completed until the threshold value reaches the upper limit value. ,
Based on the connected component information stored in the storage device, function as colony counting means for counting the number of colonies of the microorganism,
A program characterized by that.

Images