JP4213142B2 - Colony counting method - Google Patents

Description

  The present invention relates to a colony counting method in which microorganisms such as fungi and fungi are cultured in a medium using a petri dish and the number of colonies of the generated microorganisms is counted.

  Counting methods and tools for microorganisms such as bacteria and molds have been developed for medical use, and the presence or absence of bacteria and the traces that existed were mainly counted. On the other hand, in the food industry field, as a viable bacteria detection system for detecting that microorganisms subject to counting of microorganisms do not cause problems, or that useful microorganisms are growing properly in the food production process, A culture type counting means using a medium is often used. For example, as an official method for counting microorganisms, a method in which a colony of microorganisms is generated using an agar medium and the number of colonies is counted is widely used particularly for products to be counted with a low degree of biological dispersion.

Then, instead of the traditional method of visually counting the number of colonies generated on an agar medium, etc., a number of means for automatically counting the number of colonies by processing the captured image data of the medium to be counted using a CCD camera are also proposed. Has been. For example, a method of counting the number of colonies by diffusing a fluorescent substance that reacts with microorganisms on the surface of a culture medium, or a method of performing a counting process by staining a pathogen with antibody staining or the like has been proposed (Patent Document 1, Patent Document). 2). However, these methods have limitations such as time-consuming and troublesome pretreatment for colony counting, and limited target microorganisms.

  In addition, a method of counting connected colonies from the shape and counting the colonies, and a means for calculating the number of microorganisms by detecting the outer circumference arc using the shade of hue for colony microorganisms have been proposed. In addition, there was a limit in counting various colonies having large differences in shape and shape with high accuracy (see Patent Documents 3 and 4). On the other hand, while the number of colony counting systems by complex data processing using RGB hues is increasing, there is a need for means for simply processing the data of monochromatic images that do not use hues and accurately counting the number of colonies. (See Patent Document 5).

An apparatus that automatically counts the number of colonies using the analysis of these microorganisms and the colony counting method has also been proposed (see Patent Document 6 and Patent Document 7).
JP-A-7-147997 JP 2000-78999 A Japanese Patent Laid-Open No. 9-140395 JP 2001-22929 A JP 2004-194610 A JP 2001-157573 A JP 2002-98704 A

  The present invention has been proposed in view of the above-mentioned situation, and is simpler than the conventional colony counting means, and has a device configuration using a single color CCD camera, and performs complicated pretreatment for counting on the culture medium. It is an object of the present invention to provide a colony counting method that can count the number of colonies with high accuracy in a short time for various forms of microorganisms.

The invention according to claim 1 is a method of counting the number of colonies by processing data of a single color photographed image obtained by photographing a colony of microorganisms generated in a culture medium in a petri dish with a CCD camera, and photographing the photographed image A target area for capturing colonies in a petri dish as image data is set as a count image area, and the count image area is divided into a central image area and a peripheral image area, and the imaging of each area is performed. Separately count the number of colonies for image data by different data processing means, add the number of colonies in both areas to calculate the number of colonies in the counted image area, and calculate the number of colonies outside the peripheral image area The outermost peripheral area to be set is set, and the count of colonies in the counted image area is used instead of directly counting from the captured image data of the outermost peripheral area. According to the set calculation method, the number of colonies in the outermost peripheral area is calculated, and by adding this to the number of colonies in the counted image area, the number of colonies in the total colony counting area combined with the counted image area and the outermost peripheral area is calculated. The present invention relates to a colony counting method characterized by calculating .

The invention according to claim 2, in claim 1, wherein when the data processing captured image data of the central image region, colonies larger by fractional criteria set in advance, and 2 minutes in a small size, the larger size initially In the colony counting method, the colony is counted for the colony, the colonies of a small size are counted after the fact, and the number of colonies of both large and small sizes is added to count the number of colonies in the central image area. Related.

In the invention according to claim 3 , after counting the number of colonies in the count image area using the colony counting method according to claim 1 or 2 , a plurality of colonies are combined in the count image area colony. Colonies that may be present are classified as composite colonies, and colony separation data processing is performed only on the composite colonies. And the number of additional colonies isolate | separated from the composite colony is added to the number of colonies before colony separation, It concerns on the colony counting method characterized by counting the true colony number of the said count image area | region.

  According to the first aspect of the present invention, the target area for counting colonies in the single-color image of the culture medium and petri dish taken with the CCD camera is set as the count image area, and these are set as the central image area and the peripheral image area. Dividing into two, different data processing suitable for the properties of both image data is performed, and the number of colonies is calculated individually. That is, since the image characteristics and the colony distribution situation differ between the central image area and the peripheral image area, colony counting accuracy and counting efficiency can be obtained by adopting a captured image data means suitable for the image characteristics in each area. An excellent count can be performed.

In addition , the outermost peripheral area for calculating the number of colonies is set outside the peripheral image area, and the colony number of the counted image area is set in advance without directly counting from the captured image data of the outermost peripheral area. The number of colonies in the outermost peripheral area is calculated according to the calculation method. Then, by adding the calculated number of colonies in the outermost peripheral area to the number of colonies in the counted image area, the number of colonies in the total colony counting area combined with the counted image area and the outermost peripheral area is calculated. The reason for this is that at the inner peripheral edge of the petri dish, the normal photographed image is influenced by the shadow and bubbles of the petri dish, and the colony images of the petri dish vertical part overlap each other, and the colony counting accuracy is further lowered than the surrounding image area. This is to compensate for this and to ensure the colony counting accuracy.

  Based on a large number of count data collected in advance, the correlation between the number of colonies in the count image region and the number of colonies in the outermost peripheral region is preliminarily analyzed according to the type of the target microorganism and the culture conditions, and based on this correlation data A formula for calculating the number of colonies in the outermost peripheral area from the number of colonies in the count image area is set in advance. For this reason, it becomes possible to measure the colonies in the outermost peripheral area efficiently with sufficient accuracy based on the number of colonies in the counting image area according to this formula, and by adding this to the number of colonies in the counting image area The number of colonies in the entire counting area can be calculated efficiently. That is, it is possible to calculate and use practical and easy-to-use data by eliminating unnecessary fluctuation of the colony count value generated when the number of colonies in the outermost peripheral area is directly counted from the photographed image data. effective.

According to the invention described in claim 2 , when data processing is performed on the colonies in the central image area according to claim 1 , the colonies are divided into two large and small sizes according to a predetermined classification criterion, Colony counting is performed on large colonies. And the number of small colonies is counted in the state which deleted the image data regarding the colony of a large size from the data of the count area. In this case, the size of the remaining small colonies is limited to the specified size or less, and since extremely small colonies can be detected because image processing conforming to this is performed, the resulting counting accuracy can be remarkably improved. Can do.

According to the invention described in claim 3 , after counting the number of colonies in the count image area using the colony counting method according to claim 1 or 2 , first, a plurality of colonies are combined in the count image area. A colony that may be present is determined as a complex colony. And since colony separation data processing is performed only for complex colonies, the number of colonies to be separated can be reduced, and image processing means suitable for colony separation can be applied to accurately count colonies. It becomes. Then, the number of additional colonies thus separated is added to the number of colonies before colony separation, and the true number of colonies in the counting image region can be counted accurately and efficiently.

  Hereinafter, the present invention will be described in detail with reference to FIGS. FIG. 1 is a flowchart showing a colony counting process according to the method of the present invention, FIG. 2 is a side view showing an example of an apparatus for carrying out the colony counting according to the present invention, and FIG. 3 is generated with a medium in a petri dish 4 is a plan view showing the image division procedure when counting colonies, FIG. 4 is an explanatory view showing a separating means of a complex colony in which colonies are close or overlapped, and FIG. 5 is a colony at the boundary between the central image area and the peripheral image area. It is explanatory drawing when it exists.

  FIG. 1 is a flowchart showing a process of counting colonies of microorganisms such as bacteria and microorganisms cultured in a petri dish according to the method of the present invention. First, the configuration of the apparatus for counting the number of colonies will be described with reference to the side view of FIG. 2. The culture medium 1 in the petri dish 2 disposed on the diffusion plate 4 fixed to the support plate 5 constituting a part of the gantry 17 is With the petri dish lid 3 covered downward, the illumination device 6 illuminates through the parallel plate 18 from below. Above the petri dish 2, a CCD camera 7, not shown in the figure, is arranged so that a plane image of the culture medium can be taken from the bottom direction of the petri dish 2.

  Further, a signal cable 8 is connected to the CCD camera 7, the captured image data is sent to the data processing device 9, the data processing described in detail later is performed, the number of colonies K in the culture medium 1 is counted, and necessary. In response to this, the captured image or the like is displayed on the image monitor 9a. In the present embodiment, the petri dish 2 is inverted and placed on the diffusion plate 4 because, for example, if the petri dish lid 3 is removed during imaging of the medium, it becomes difficult to continue culturing with various bacteria entering the medium, The illumination device 6 illuminates and the CCD camera 7 photographs from above. Note that the illumination device 6 and the CCD camera 7 can be arranged in reverse, and are not particularly limited to the illustrated embodiment.

  FIG. 3 shows an example of a photographed image 14 obtained by photographing the bottom surface of the petri dish 2 with the upper CCD camera 7. The culture medium 1 is inside the petri dish side wall 2a, and the colonies K of microorganisms to be counted are generated in various shapes. Has been. Here, the image inside the petri dish side wall 2a is divided into a central image region 10, a peripheral image region 11, and the outermost region 12 as shown in the figure. Note that a region obtained by combining the central image region 10 and the peripheral image region 11 is defined as a count image region 13 in which the number of colonies is counted by data processing of photographed image data.

  At this time, the size or ratio of the central image region 10, the peripheral image region 11, and the outermost peripheral region 12 with respect to the inner diameter of the petri dish side wall 2a is not particularly specified, but the central image region 10 has a distribution of colonies K based on past experience. In addition, an area that is optically less affected by the petri dish side wall 2a and that can obtain a relatively good image is selected, and the peripheral image area 11 is affected by the distribution of the colonies K and the optical influence of the petri dish side wall 2a for image processing. An area with some difficulty is designated, and the outermost peripheral area 12 designates an area that is extremely difficult to count the colonies K by the photographed image in the vicinity of the petri dish side wall 2a. For example, when the diameter of the petri dish inner wall 2a is 100, the diameter of each of the central image region 10, the peripheral image region 11, and the outermost peripheral region 12 is set to 84, 92, 100, and the number of colonies is counted. To do.

  The colony counting process and the image data processing means will be described below with reference to the flowchart shown in FIG. 1 and FIG. First, in the photographed image data input (S1), the petri dish 2 and the internal culture medium 1 are photographed by the CCD camera 7, and the photographed image data of the photographed image 14 is transmitted to the data processing device 9 by the signal cable 8 and input. Next, in the count image generation (S2), an unnecessary image outside the petri dish 2 is obtained by means such as detecting the position of the petri dish 2 in the photographed image, obtaining the center of the petri dish, determining alignment, and image size. Masking and erasing are performed to generate an image to be counted corresponding to FIG.

  In the counting area automatic extraction (S3), when the inner diameter of the petri dish side wall 2a is set to 100, for example, the diameter of the counting image area 13 is set to a predetermined ratio such as 92 as illustrated above, and the counting image area 13 Is automatically extracted and determined. The diameter ratio of the count image region 13 to the diameter of the petri dish side wall 2a specified in advance is set to a suitable ratio based on past count data in consideration of the type of microorganism to be counted, culture conditions, the size of the petri dish 2, and the like. .

  Next, in the division of the measurement area (S4), the count image area 13 is divided into two, the central image area 10 and the peripheral image area 11, and the colonies in each area are counted individually (S5, S6). This is because the photographed image in the vicinity of the petri dish side wall 2a is affected by the shadow of the petri dish 2, bubbles in the culture medium, and the plastic material injection gate when the petri dish 2 is molded. Comparing to the fact that there are more shade fluctuations and a lot of noise and data processing accuracy is reduced, image data pre-processing and counting means suitable for each area are adopted, colony counting accuracy and counting efficiency It is intended to increase.

  In general, when a colony medium image is used as a target, the density distribution is in a narrower range than a general natural image, so that the contrast is low and it is often difficult to identify an inspection object. For this reason, in the embodiment of the present invention, the density value is converted to use the entire density value area within the allowable range of the display function of the data processing device 9 with respect to the entire counting image area 13, and the density dynamic range. Is expanded to facilitate the detection and counting of the colony K, and multi-value screen data is preprocessed.

  Thereafter, in the counting of the peripheral image area (S5), for example, a Laplace filter is applied to the photographing surface image data of the peripheral image area 11 as spatial filtering to extract the contour of the colony K, and then the binarized image is extracted. create. Then, after removing noise from the binarized image data without blurring, the connected components are labeled and colony K is counted.

  At this time, the minimum value of the measurement area of the connected component and the maximum value of the measurement major axis are set as colony identification threshold values, the measurement area is larger than the set minimum value, and the measurement major axis is within the set ratio range with respect to the measurement area. Those belonging to the genus and having characteristics as colonies are judged as colonies K and counted. Note that the preprocessing of image data as the colony counting means and noise removal as the above-described embodiment are not particularly limited to the exemplified means, and various data processing means can be used according to the characteristics of the target captured image. It can be selected and applied.

  Next, in the counting of the central image area (S6), the binarization is performed on the central image area 10 of the multi-value screen data in which the density dynamic range is expanded by the discriminant analysis method of the automatic threshold value determination method. The method is adopted. That is, for the peripheral image region 11, a binary image is created by adopting one threshold for the whole, but it occupies the main part of the area of the counting region 13 and has a large influence on the colony counting accuracy. For the central image region 10, the threshold value is determined so that the difference between the colony K as the object and the background medium is distinguished using the discriminant analysis method of the automatic threshold value determination method, and the binarized image is displayed. The connected components of the binarized image are labeled, and colonies K are counted using a preset area and shape threshold for the colonies.

  In the above-described counting of the central image area (S6), the colony counting of the central image area was performed together, but the counting of the central image area was performed by extracting and counting large colonies (S7) and extracting and counting small colonies ( As in S8), the colony K in the central image region can be separated into large and small colonies and counted. In this case, first, as shown in the extraction and counting (S7) of the large colony K, the central image area 10 is subjected to the determination of the colony K as the object and the background using the discriminant analysis method of the automatic threshold determination method. Create a binarized image by determining the threshold value so that the difference from the culture medium stands out. Find the connected components of the binarized image, and use the area and shape threshold values for the large colony set in advance. Count.

  Next, the large colony image in the center plane image area 10 is masked, and the captured image data is binarized by the variable threshold method for the remaining area, and the small colony K is selected and counted. (S8). That is, depending on the position in the petri dish 2, there may be a portion where the density of the photographing screen or the density of the colony itself is locally high or low, so that the corresponding pixel is binarized by comparing with the local average of neighboring pixels. A binarized image is created by the variable threshold method, and even a very small colony K is surely revealed as an image so that it can be accurately counted.

  In this case, for the central image region 10 having a high area ratio and a large influence on the colony counting accuracy, the colonies K in the region are sorted and counted by an image processing means suitable for counting large and small colonies K (S7, After S8), the count values are added at the stage of the primary integration (S9), thereby making it possible to perform highly accurate counting for the extremely small colony K. Note that the pre-processing of the photographed image data and the threshold value determination means in the colony counting are various in accordance with the characteristics of the target microorganism, the generation conditions, and the characteristics of the CCD camera 7 and the lighting fixture used for the colony counting. These means can be selectively employed and are not particularly limited to the embodiments.

  In the next integration of the count values (S9), the number of colonies in the central image area 10 and the peripheral image area 11 is added to obtain the number of primary colonies in the count image area 13. Thereafter, in the determination of the separation target (S10), there is a possibility that a large colony K may include a plurality of colonies K with the area of the colony K designated in advance from the colonies of the count image region 13 as a threshold value. Sort as a colony K. That is, the colonies K1 to K4 shown in the plan view of FIG. 4 were processed and counted as one colony K in the processing of the image data until the primary integration (S9) of the count values.

  And as a separation algorithm (S11) which isolate | separates the compound colony image K1 thru | or K4 as an Example into each individual colony K, it is with the smoothing operator of Gauss, for example with respect to the picked-up image data of the colony K made into object. Neighboring colonies are separated by using a light and dark boundary, and a mountain having a high concentration is separated as a colony K for overlapping colonies. As a result, the colony K1 is separated into a plurality of colonies K such as K1a, K1b, the colony K2 is K2a, K2b, the colony K3 is K3a, K3b, K3c, and the colony K4 is K4a, K4b, K4c. In addition, the colony K3b and the colony K4c in the colonies K1 to K4 are overlapping colonies, and others are neighboring colonies, and the overlapping colonies are shown when colonies K are generated at different heights in the thickness direction in the medium.

  As described above, the separation algorithm (S11) is used to separate the individual colonies constituting the composite colony, and the resulting increased number of colonies is added in the secondary integration (S12) of the count values to obtain the central image. The true colony numbers of the area 10 and the peripheral image area 11 are obtained. Then, in the necessity of correction of the outermost peripheral area 12 (S13), it is determined whether it is necessary to correct by adding the number of colonies in the area on the petri dish side wall 2a on the outer periphery of the peripheral image area 11 as management data to be used. In addition, when using the counted number of colonies as management data, when additionally using the number of colonies in the outermost peripheral region 12, the outermost peripheral region 12 is placed outside the peripheral image region 11 in the automatic correction by area (S14). And the number of colonies near the petri dish side wall 2a is calculated as follows using the counted number of colonies in the central image region 10 and the peripheral image region 11 (S14).

  For example, the area of the set outermost peripheral region 12 is calculated, and the number of colonies in the outermost peripheral region 12 is obtained by multiplying this by the number of colonies in consideration of the composite colony per area of the peripheral image region 11. In addition, the ratio of the number of colonies in the outermost peripheral region 12 to the number of colonies per area of the peripheral image region 11 is obtained in advance based on the count test data, and the number of colonies in the outermost peripheral region 12 is calculated based on this ratio. The number estimation accuracy can be further improved.

  The reason why the number of colonies in the outermost peripheral region 12 is not directly counted from the captured image data and is calculated from the number of colonies in the central image region 10 and the peripheral image region 11 is that the captured image of the region close to the petri dish side wall 2a is The influence of petri dish shadows and bubbles and the colony images in the vertical part of the petri dish overlap each other and the colony counting accuracy is greatly reduced. This is because more effective results are obtained in counting and managing the colony K.

  As shown in FIG. 5, for the colonies K13 to K15 on the boundary of the central image region 10, the peripheral image region 11, and the outermost peripheral region 12, for example, the colony K is counted by the area ratio of the colonies K of each region, Convenient operation such as using the number of colonies rounded off to the last 5 can be performed. Then, the final number of colony counts can be obtained by summing the number of colonies in the outermost peripheral area 12 and the count image area 13 (S15). Moreover, the data of the colony counting result can be stored in the data processing device 9 as needed, processed into a useful form, and used as management data.

  In the above description, specific examples have been described regarding the data processing of the captured image and the colony K counting method for each of the examination areas. However, the present invention is not limited to the scope of the examples, and various kinds of embodiments are within the scope of the present invention. Data processing means and counting methods can be employed. Moreover, although this invention demonstrated the count method of the colony K in the culture medium 1 using the petri dish 2, it replaces with the petri dish 2 and applies also to culture media of other systems, such as the colony produced | generated to the culture medium using a plastic sheet. Can do.

  Further, in the embodiment, as a means for removing noise from the image data, the photographed image is divided into a central image region and a peripheral image region, and an image of the peripheral image region that includes a lot of noise variation and a lot of noise as compared with the central image region. A method for removing binarized image data from data and using noise is used. However, after picking up a plurality of picked-up images and comparing the plurality of picked-up images, the noise is specified, and the noise is processed electronically. The obtained image data may be used for the counting process.

It is a flowchart which shows the process of the colony count which concerns on the method of this invention. It is a side view which shows the example of the apparatus for implementing the colony count which concerns on this invention. It is a top view which shows the example of a division | segmentation at the time of the count of the colony in a petri dish. It is explanatory drawing which shows the isolation | separation means of the composite colony in which a colony adjoins or has overlapped. It is explanatory drawing of the counting method when a colony exists in the boundary of a center image area | region and a periphery image area | region.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Medium 2 Petri dish 2a Petri dish side wall 3 Petri dish lid 4 Diffusion plate 5 Support plate 6 Illumination device 7 CCD camera 8 Signal cable 9 Data processing device 9a Image monitor 10 Central image area 11 Peripheral image area 12 Outermost circumference area 13 Counting image area 14 Photographing Image 17 Mount 18 Parallel plate K Colony K1 Colony K2 Colony K3 Colony K4 Colony K13 Colony K14 Colony K15 Colony S1 Captured image data input S2 Counting image generation S3 Counting area automatic extraction S4 Counting of surrounding image area S6 Center image S6 Center image correction necessity S of the secondary integration S13 outermost region of the primary integrated S10 separation target determination S11 separation algorithm S12 the count value of the extraction and counting S9 the count value of the counting S8 small colonies extraction counting S7 large colony area The final value of the automatic correction S15 counting by 4 area

Claims (3)

It is a method of counting the number of colonies by processing data of a single color image obtained by photographing a colony of microorganisms generated in a medium in a petri dish with a CCD camera,
Taking the captured image as captured image data into a data processing device and setting a target area for counting colonies in a petri dish as a count image area,
The counting image area is divided into a central image area and a peripheral image area, the number of colonies is individually counted by different data processing means for the captured image data of each area, and the number of colonies in both areas is added. While calculating the number of colonies in the counting image area ,
A preset calculation method that sets the outermost peripheral area for calculating the number of colonies outside the peripheral image area, and uses the counted colony number of the counted image area without directly counting from the captured image data of the outermost peripheral area. Calculating the number of colonies in the outermost peripheral area according to the above, and calculating the number of colonies in the total colony counting area combined with the counted image area and the outermost peripheral area by adding this to the number of colonies in the counted image area. Characteristic colony counting method. In claim 1 , when processing the captured image data of the central image region, the colony is divided into large and small sizes according to a preset classification criterion, and colonies are counted for large colonies first, A colony counting method comprising: counting colonies of a small size after the fact, and adding both to count the number of colonies in the central image region. After counting the number of colonies in the count image area using the colony counting method according to claim 1 or 2 , a colony in which a plurality of colonies may be complexed among the colonies in the count image area is a complex colony. The colony separation data processing is performed only on the composite colony, the number of additional colonies separated is added to the number of colonies before colony separation, and the number of true colonies in the counting image area is counted. A colony counting method characterized by the above.

Images