JP2015171333A - Determination method, determination program, and determination device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To evaluate the quality of check.SOLUTION: A server device 10 detects positions of colonies of bacteria cultivated in a Petri dish with respect to each of a plurality of colonies by performing an image processing of the captured images of the Petri dish where the bacteria is cultivated. Furthermore, the server device 10, on the basis of each position of the detected plurality of colonies, calculates planar distribution conditions of the plurality of colonies to determine whether the calculated distribution conditions satisfy prescribed distribution conditions, and outputs the results of the determination.

Description

  The present invention relates to a determination method, a determination program, and a determination apparatus.

  In food production facilities, etc., hygiene inspections are performed to inspect for contamination of fungi. For example, a specimen from which a part of the food is collected is smeared or mixed in a medium prepared on a petri dish, and bacteria contained in the specimen are kept at an appropriate temperature on the petri dish and cultured for a predetermined period. Thereafter, the hygiene inspection is performed by counting the colonies formed on the petri dish by the culture.

  As an example of a technique for supporting such a hygiene inspection, a technique for counting colonies from an image obtained by imaging a colony cultured on a petri dish has been proposed.

JP 2013-21960 A JP 2003-85533 A

  However, the above technique cannot evaluate the inspection quality as described below.

  That is, when the above-described sanitary inspection is performed, so-called contamination may occur as one of accidents in the culture process. For example, when the specimen is smeared on the culture medium on the petri dish, the operator's fingers may touch the culture medium or a specimen other than the specimen to be sanitized may be mixed. As a result, other bacteria of a different type from the bacteria originally contained in the specimen may enter and be cultured on the medium. Even if the above-mentioned contamination does not occur, the inspection procedure may be deficient. For example, if the specimen smeared on the petri dish medium is uneven, the colonies grow densely in the part painted thicker than the other parts, so that the colonies are far enough to distinguish each colony by image processing In some cases, they may adhere. Even if the inspection is carried out using the petri dish in which an accident or deficiency has occurred in this way, the result of the inspection is not reliable and the inspection quality is impaired.

  However, the technique described above is merely a count of colonies, and even if the colony count result is used, it is difficult to evaluate the inspection quality. For example, it cannot be determined that a sanitary inspection is being carried out without hindrance simply because colony count results are equivalent to petri dishes that were free of accidents and inadequacies. This is because just because the number of colonies is the same, the shape and size of the colonies are not necessarily similar. In addition, when there is unevenness in the method of applying the specimen, multiple colonies that adhere to each other at the place where the specimen is applied thicker than the other parts on the petri dish culture surface are not individually detected by image processing and are counted. This is because the result itself may not be reliable. For these reasons, the inspection quality cannot be evaluated with the above technique.

  In one aspect, an object of the present invention is to provide a determination method, a determination program, and a determination apparatus that can evaluate inspection quality.

  In one aspect of the determination method, a computer executes a process of detecting, for each of a plurality of colonies, a position of a bacterial colony cultured in the petri dish by performing image processing on a captured image of the petri dish in which bacteria are cultured. . Further, the computer calculates a planar distribution state of the plurality of colonies based on the detected positions of the plurality of colonies, and whether or not the calculated distribution state satisfies a predetermined distribution condition. A determination is performed, and a process of outputting the determination result is executed.

  The inspection quality can be evaluated.

FIG. 1 is a diagram illustrating the configuration of the inspection support system according to the first embodiment. FIG. 2 is a block diagram illustrating a functional configuration of the server apparatus according to the first embodiment. FIG. 3 is a diagram illustrating an example of a division method. FIG. 4 is a diagram illustrating an example of a frequency distribution to be evaluated. FIG. 5 is a diagram illustrating an example of the reference frequency distribution. FIG. 6 is a flowchart illustrating the procedure of the colony counting process according to the first embodiment. FIG. 7 is a flowchart illustrating the procedure of the determination process according to the first embodiment. FIG. 8 is a diagram illustrating an application example of the dividing method. FIG. 9 is a diagram illustrating an example of a frequency distribution to be evaluated. FIG. 10 is a diagram showing a columnar graph of the frequency distribution shown in FIG. FIG. 11 is a diagram illustrating an application example of the division method. FIG. 12 is a diagram illustrating an example of a frequency distribution to be evaluated. FIG. 13 is a graph showing an example of the relationship between the number of colonies and the degree of similarity. FIG. 14 is a diagram illustrating an example of a frequency distribution of similarities. FIG. 15 is a schematic diagram illustrating an example of a computer that executes a determination program according to the first and second embodiments.

  Hereinafter, a determination method, a determination program, and a determination apparatus according to the present application will be described with reference to the accompanying drawings. Note that this embodiment does not limit the disclosed technology. Each embodiment can be appropriately combined within a range in which processing contents are not contradictory.

[System configuration]
FIG. 1 is a diagram illustrating the configuration of the inspection support system according to the first embodiment. An inspection support system 1 shown in FIG. 1 counts colonies in a petri dish imaged from petri dishes in which bacterial colonies were cultured from food samples shipped from bases 3A to 3C such as factories and sales offices. . Accordingly, the inspection support system 1 provides an inspection support service that supports the hygiene inspection. As part of this inspection support service, the inspection support system 1 also provides a quality management service that supports the quality control of the hygiene inspection performed at each of the sites 3A to 3C.

  As shown in FIG. 1, the inspection support system 1 accommodates a server device 10, client terminals 30 </ b> A to 30 </ b> C, and a head office terminal 50. Although FIG. 1 illustrates three client terminals and one head office terminal, the number of devices accommodated in the system is not limited to the illustrated configuration. That is, the inspection support system 1 can accommodate any number of client terminals and head office terminals. In the following description, the bases 3A to 3C are collectively referred to as “base 3” and the client terminals 30A to 30C are collectively referred to as “client terminal 30”. There is a case.

  The server device 10, the client terminal 30, and the head office terminal 50 are connected via the network 7 so as to communicate with each other. The network 7 can employ any type of communication network such as the Internet, LAN (Local Area Network), and VPN (Virtual Private Network) regardless of wired or wireless.

  Among these, the server device 10 is a computer that provides the inspection support service and the quality management service to the client terminal 30 and the head office terminal 50. As an embodiment, the server device 10 may be implemented as a Web server that realizes an inspection support service or a quality management service, or is implemented as a cloud that provides an inspection support service or a quality management service by outsourcing. You can also. As another embodiment, it can be implemented by preinstalling or installing a program that realizes an inspection support service or a quality control service as package software or online software in a desired computer.

  The client terminal 30 is a computer that receives provision of the inspection support service and the quality management service. This client terminal 30 is used as an example by a person concerned in the inspection department at the base 3.

  As an aspect of the client terminal 30, an imaging device itself that can image a petri dish used for a hygiene inspection, a petri dish acquired via a computer, an imaging device, or an auxiliary storage device that is built-in and attached or connected by communication. A computer that can store images can be used. For example, as an example of the above-described imaging apparatus, a dedicated imaging apparatus that can capture a petri dish image while changing the distance and direction for capturing the petri dish in which the colony is cultured, and further, the type of illumination applied to the petri dish is employed. be able to. As another example, a general-purpose computer equipped with an imaging device as a standard can be used. For example, in addition to fixed terminals such as personal computers, smartphones, mobile phones, mobile communication terminals such as PHS and PDA, and tablet terminals that do not have the ability to connect to a mobile communication network can be employed. . The PHS is an abbreviation for “Personal Handyphone System”, and the PDA is an abbreviation for “Personal Digital Assistants”.

  Here, an example of the flow of sanitary inspection will be described. For example, the hygiene inspection is performed by an inspector who belongs to the inspection department of the base 3. Such an inspector collects a part of the food as the object of the hygiene inspection as a specimen and cultures bacterial colonies in the petri dish by the following method.

  For example, in a sanitary inspection, a part of food is collected as a sample by an inspector, and sample identification information, for example, a sample number assigned to each sample is given to the collected sample. Next, the inspector puts the specimen in a petri dish together with the dissolved medium in the case of the pour method, for example, in accordance with various test items such as general live bacteria, Escherichia coli and Staphylococcus aureus. At this time, it is also possible to create a plurality of petri dishes in which the sample is diluted at the same stage for the same sample. In addition, the same specimen number is written on the petri dish lid so that the test items to be performed between the same specimens can be discriminated in the petri dishes in which the same specimen is cultured. In addition, the inspector cultivates the bacteria in the petri dish by keeping the petri dish in an incubator for a predetermined period according to the inspection item, for example, about 48 hours if it is a general living microbe. A part of the sample collected from the food can be stored frozen in preparation for retesting.

  After the culture is performed in this manner, the inspector places the petri dish at a predetermined photographing position to cause the imaging apparatus to pick up the petri dish image. When such a petri dish image is captured, the petri dish image is uploaded to the server device 10 with a predetermined opportunity together with the sample number and the like. For example, uploading may be executed every time a petri dish image is taken, or uploading may be executed at a certain cycle or regular time. Thereafter, the inspector specifies the sample number for which the count result is desired to be browsed or requests the petri dish list display via the client terminal 30, whereby the colony count result counted by the server device 10 is obtained. Can be displayed on the client terminal 30. Thereby, the inspector can carry out the hygiene inspection by judging OK or NG from the colony counting result, that is, the presence or absence of colonies or the number of colonies. At this time, the determination result of the inspection quality determined by the server device 10 can also be displayed on the client terminal 30 together.

  The head office terminal 50 is a computer that receives the above-described quality management service. As an example, the head office terminal 50 is used by a person concerned in the quality control department of the head office 5.

  As one aspect of the head office terminal 50, in addition to a fixed terminal such as a personal computer, a smartphone, a mobile phone, a mobile communication terminal such as a PHS and a PDA, and further a capability to connect to a mobile communication network. You can use tablet devices that you don't have. Here, it is assumed that the quality control service is provided to the persons related to the quality control department of the head office 5 through the head office terminal 50, but the persons related to the inspection department of each base 3 such as the person in charge of the inspection department or the like. The quality management service may be provided through the client terminal 30 used by the parties concerned. In this case, the person in charge at the base 3 or the person concerned can browse only the sanitary inspection related to each base 3.

  In the following, a petri dish image is illustrated as an example of a captured image. However, a bacterial colony may be included in the captured image, and the petri dish may not be captured in an environment where the bacterial colony exists. In addition, in the following, as an example, the following explanation will be given by exemplifying a scene in which the quality control department of the head office 5 investigates the quality of the hygiene inspection performed in the inspection departments of the respective bases 3A to 3C. Is not limited to this example.

[Configuration of Server Device 10]
FIG. 2 is a block diagram illustrating a functional configuration of the server apparatus 10 according to the first embodiment. As illustrated in FIG. 2, the server device 10 includes a communication I / F (InterFace) unit 11, a storage unit 13, and a control unit 15. The server device 10 may include various functional units included in a known computer other than the functional units illustrated in FIG. 2, for example, functional units such as various input / output devices, audio output devices, and imaging devices. Absent.

  Among these, the communication I / F unit 11 is an interface that performs communication control with other devices, for example, the client terminal 30 and the head office terminal 50. As an aspect of the communication I / F unit 11, a network interface card such as a LAN card can be employed. For example, the communication I / F unit 11 receives a petri dish image from the client terminal 30 or transmits a colony count result to the client terminal 30. In addition, the communication I / F unit 11 receives an inspection quality browsing request from the head office terminal 50, and transmits the inspection quality determination result to the head office terminal 50.

  The storage unit 13 is a storage device that stores various programs such as an OS (Operating System) executed by the control unit 15, a program for realizing an inspection support service and a quality management service. As one aspect of the storage unit 13, a storage device such as a semiconductor memory element such as a flash memory, a hard disk, or an optical disk can be cited. The storage unit 13 is not limited to the type of storage device described above, and may be a RAM (Random Access Memory) or a ROM (Read Only Memory).

  The storage unit 13 stores inspection data 13a as an example of data used in a program executed by the control unit 15. In addition to the inspection data 13a, other electronic data, for example, inspection items for a hygiene inspection performed at the base 3, inspectors performing the inspection items, and sample numbers of petri dishes used for the inspection items are associated with each other. Implementation data and the like can also be stored. The inspection data 13a includes intermediate data generated in the process of execution by the control unit 15 and will be described together with the description of the control unit 15.

  The control unit 15 has an internal memory for storing programs defining various processing procedures and control data, and executes various processes using these. As shown in FIG. 2, the control unit 15 includes a receiving unit 15a, a detecting unit 15b, a counting unit 15c, a dividing unit 15d, a calculating unit 15e, a determining unit 15f, and an output unit 15g.

  Among these, the reception unit 15 a is a processing unit that receives a petri dish image from the client terminal 30. As an embodiment, the reception unit 15 a can set the client terminal 30 at a timing for uploading the petri dish from the client terminal 30 to the server device 10. For example, the reception unit 15a can upload the petri dish in real time every time the petri dish is captured. In addition, when the upload instruction operation is executed on the client terminal 30, the reception unit 15a can upload the petri dish that has been stored in the client terminal 30 until then. Further, the reception unit 15a may upload the petri dish at a certain period, for example, every 30 minutes, one hour or one day, or may upload the petri dish at a regular time, for example, 12:00 or 18:00. It doesn't matter. The accepting unit 15a can also register the petri dish image received from the client terminal 30 in the storage unit 13 before the colony counting is executed by the subsequent counting unit 15c.

  The detection unit 15b is a processing unit that detects colonies from the petri image.

  As one embodiment, the detection unit 15 b generates a histogram of luminance values by referring to the luminance values of each pixel included in the petri dish received from the client terminal 30. Then, the detection unit 15b calculates the luminance value having the highest frequency in the luminance value histogram as the luminance value representing the background portion such as the culture medium. After that, the detection unit 15b calculates a threshold value to be compared with the luminance difference between the luminance value of the pixel in the petri image and the representative luminance value of the background portion. In this way, the representative luminance value and threshold value of the background portion of the petri image are derived as colony detection criteria. Here, the case where the luminance value having the highest frequency in the histogram of luminance values is used as the luminance difference between the representative luminance values of the background portion is exemplified, but other statistical values such as an average value and a median value are used. It is good.

  After the colony detection criterion is derived in this way, the detection unit 15b generates the binarized image from the petri dish image according to the colony detection criterion. As an example, the detection unit 15b determines whether or not the luminance difference between the luminance value of the pixel in the petri dish image and the representative luminance value of the background portion is greater than or equal to a threshold value. At this time, if the luminance difference is less than the threshold, it can be considered that the pixel corresponds to the background. In this case, the detection unit 15b sets the pixel value of the pixel to “0”, that is, “black”. On the other hand, when the luminance difference is equal to or greater than the threshold value, it is unlikely that the pixel corresponds to the background, so that it can be estimated as a part of the colony. In this case, the detection unit 15b sets the pixel value of the pixel to “1”, that is, “white”. Thus, a binarized image can be obtained by executing the above binarization for each pixel in the petri image.

  Thereafter, the detection unit 15b estimates the presence of a colony included in the binarized image from the binarized image generated previously. For example, the detection unit 15b performs edge detection on the binarized image. And the detection part 15b labels the edge or group of edges which forms the outline of the shape similar to the shape which the colony made into a test | inspection has among the edges obtained by previous edge detection as a colony candidate. Then, the detection unit 15b detects a colony candidate similar to the size of the type of colony to be inspected as a colony among the colony candidates. As an example, the detection unit 15b determines a colony candidate as a colony when the number of pixels included in the colony candidate is within a predetermined appropriate range and an area defined by the type of colony to be inspected. Can be detected. In addition, the detection unit 15b stores the colony position where the above-described detection is performed, for example, the coordinates of the center of gravity position in an internal memory (not shown).

  The counting unit 15c is a processing unit that counts colonies. As one embodiment, the counting unit 15c counts the total number of colonies detected by the detecting unit 15b from the binarized image of the petri image. After that, the counting unit 15c stores the count result of the colonies counted in this way as a test data 13a after associating it with a petri dish image, a binarized image, a specimen number, a position of each colony, and the like. 13 is registered.

  The dividing unit 15d is a processing unit that divides the petri image into a plurality of regions. As one embodiment, the dividing unit 15 d activates the process when receiving an inspection quality browsing request from the head office terminal 50. In addition, the dividing unit 15d can also start the process when the counting result by the counting unit 15c is obtained, and can also target the inspection data 13a whose inspection quality is undetermined at a certain period or regular time. You can also start the process. For example, the dividing unit 15d reads the inspection data corresponding to the inspection number for which the inspection quality browsing request has been received from the head office terminal 50 among the inspection data 13a stored in the storage unit 13. In addition, the dividing unit 15d divides the petri dish included in the inspection data read out previously into areas partitioned into a lattice shape, a so-called mesh shape. For example, the dividing unit 15d divides the petri dish image into a predetermined number, in this example, an area divided into 8 × 8 meshes as an example. In other words, as an example, the dividing unit 15d divides the petri dish into 64 blocks having the same size.

  FIG. 3 is a diagram illustrating an example of a division method. In FIG. 3, the illustration of the colony is omitted from the viewpoint of block visibility, but the actual image may include the colony. As shown in FIG. 3, the petri dish image is divided into 64 blocks by setting an 8 × 8 mesh in the petri dish portion shown in the petri dish image. Here, among the blocks included in the petri dish image, a block number for identifying a block is given by focusing on a region of interest, which is a target of processing executed by a functional unit at a later stage, a so-called ROI (Region of Interest). Specifically, as shown in FIG. 3, when the petri dish is circular, the block includes a region that does not contain the petri dish. Avoiding such blocks, block numbers are given only to blocks that show petri dishes. In the example of FIG. 3, the block numbers “1” to “52” are assigned to the 52 blocks illustrated. In addition, although the case where the block number is assigned to the block in which the petri dish is shown is illustrated here, the block number may be assigned to each block.

  The calculation unit 15e is a processing unit that calculates the planar distribution status of the plurality of colonies based on the positions of the plurality of colonies.

  As one embodiment, the calculation unit 15e refers to the position of each colony included in the inspection data and calculates the frequency distribution of the number of colonies. For example, the calculation unit 15e calculates the frequency distribution of the number of colonies by counting the number of colonies present in each region divided by the division unit 15d. The frequency distribution of the number of colonies calculated in this way means the distribution situation of colonies for each region where the culture surface of the petri dish is divided. After calculating the frequency distribution of the number of colonies, the calculation unit 15e further derives the frequency distribution of the reference colony number to be compared with the frequency distribution of the colony number of the petri dish that is the object of the inspection quality evaluation. In the following, the frequency distribution of the number of colonies in the petri dish, which is the evaluation object of the inspection quality, will be described as “frequency distribution of the evaluation target”, while the frequency distribution of the reference colony number will be referred to as “reference frequency distribution”. May be described.

  Here, as an example of the “reference frequency distribution”, a frequency distribution in which the frequencies are equal or substantially equal between the regions, in other words, a frequency distribution in which colonies are uniformly distributed on the petri dish image is adopted. As an example, the calculation unit 15e calculates a reference frequency distribution having a frequency in which total frequencies similar to the total frequency of the “frequency distribution to be evaluated” are evenly allocated to each region. Specifically, the calculation unit 15e divides the total frequency of the “frequency distribution to be evaluated” by the number of regions. The division value obtained by such division is calculated as the frequency of each region of the “reference frequency distribution”.

  FIG. 4 is a diagram illustrating an example of a frequency distribution to be evaluated, and FIG. 5 is a diagram illustrating an example of a reference frequency distribution. The vertical axis of the columnar graphs shown in FIGS. 4 and 5 indicates the frequency of the number of colonies, and the horizontal axis of the columnar graphs indicates the block number. In FIG. 4, the number of colonies distributed in the block number “1” to block number “52” shown in FIG. 3 is displayed as the frequency, and the total frequency, that is, the colony count result is displayed. Is “244”. When the frequency distribution shown in FIG. 4 is obtained as “frequency distribution to be evaluated”, as shown in FIG. 5, the total frequency “244” is divided by the total number of ROI blocks “52”. 4.7 (≈4.6923...) ”Is derived as the frequency corresponding to each block of the reference frequency distribution. This reference frequency distribution means that 4.7 colonies are uniformly distributed in each block of block number “1” to block number “52”. Here, the case where the reference frequency distribution is calculated from the frequency distribution to be evaluated has been illustrated, but a reference frequency distribution for each absolute value of the total frequency may be registered in advance.

  The determination unit 15f is a processing unit that determines whether the distribution state satisfies a predetermined distribution condition.

As one embodiment, the determination unit 15f calculates the similarity between the frequency distribution and the reference frequency distribution of the evaluation target calculated by the calculation unit 15e. For example, a histogram intersection shown in the following equation (1) can be used to calculate the similarity. In the following equation (1), “H ₁ ” indicates the frequency distribution to be evaluated, “H ₂ ” indicates the reference frequency distribution, and “N” indicates the total number of classes, that is, the total number of blocks. . When using this Histogram Intersection, determination unit 15f has a minimum value min among the frequency with the reference frequency distributions H ₂ in degrees and the class i with the frequency distribution H ₁ to be evaluated in the class i for each class i The sum Σ of the minimum values min of each class 1 to N is selected. On top of that, the determination unit 15f may calculate the similarity to take ~ 1 the following values 0 or more by dividing a minimum value total frequency of the frequency distribution H ₁ the sum Σ of the evaluated min for each class 1~N it can. The closer the degree of similarity is, the closer the frequency distribution of both is, and the closer the value is to 0, the less similar the frequency distribution of both. For example, when the degree of similarity is 1, the frequency distribution of both Means match.

For example, a calculation example of similarity will be described using the examples of FIGS. In the case of the class 1, that is, the block with the block number “1”, the frequency of the evaluation target frequency distribution H ₁ is “5”, and the frequency of the reference frequency distribution H ₂ is “4.7”. For this reason, “4.7” is selected as the minimum value of class 1. Further, in the case of the class 5, that is, the block with the block number “5”, the frequency distribution of the frequency distribution H _{1 to} be evaluated is “4”, and the frequency of the reference frequency distribution H ₂ is “4.7”. For this reason, “4” is selected as the minimum value of class 1. Hereinafter, although explanation is omitted, the minimum value min of each of the classes 1 to 52 is selected in the same manner. In this way, the sum Σ of the minimum values min of the classes 1 to N is calculated as “221”. Then, the sum Σ “221” of the minimum values min of the classes 1 to N is divided by the total frequency “244” of the frequency distribution H _{1 to} be evaluated, whereby the similarity is “0.91≈0.9057. ... In addition, although the case where the frequency of each block was calculated equally was illustrated here, the frequency of each block can also be normalized. For example, in some of the petri dish images shown in FIG. 3, for example, blocks such as block numbers “1” and “4” contain only a part of the petri dish. Normalization can also be performed by dividing by the proportion of the area of the petri dish occupying.

  After the similarity is calculated in this way, the determination unit 15f determines whether or not the similarity between both frequency distributions is equal to or greater than a predetermined threshold. As this threshold value, an arbitrary value from 0 to 1 can be adopted, but as an example, the value can be set based on the colony count result, that is, the total frequency of the frequency distribution of the evaluation target. it can. For example, a smaller value can be set as the threshold value as the number of colonies included in the petri dish is smaller, while a larger value can be set as the threshold value as the number of colonies is larger.

  Here, when the similarity is equal to or greater than a predetermined threshold, it is found that the frequency distribution to be evaluated and the reference frequency distribution are similar. Among these, since the standard frequency distribution has a uniform colony distribution in each region, the frequency distribution to be evaluated similar to this is equivalent to a uniform colony distribution in each region. In this case, it can be presumed that the specimen is uniformly applied to the petri dish medium shown in the petri image, and that there is little uneven coating of the specimen. In addition, when colonies are uniformly distributed in each region, it is unlikely that colonies of different sizes are growing on the petri dish, so other types of samples different from the type of sample to be examined are mixed. It is estimated that there is a low possibility that From these facts, when the similarity is equal to or higher than the threshold, it can be considered that there is a low possibility that contamination has occurred or that there is a defect in the inspection procedure such as uneven coating of the specimen. .

  On the other hand, when the similarity is less than the threshold, it is found that the frequency distribution to be evaluated and the reference frequency distribution are not similar. Among these, since the standard frequency distribution has a uniform colony distribution in each region, the frequency distribution to be evaluated that is not similar to this is equivalent to the non-uniform colony distribution in each region. In this case, it can be estimated that the specimen applied to the petri dish medium shown in the petri dish image is not uniform, and there is a high possibility that the specimen is unevenly coated. Furthermore, if colonies are not uniformly distributed in each region, colonies with different sizes may be growing on the petri dish, so other types of samples different from the type of sample to be examined It can be estimated that there is a high possibility of contamination. From these facts, when the similarity is less than the threshold value, it can be considered that there is a high possibility that contamination has occurred or that there is a defect in the inspection procedure such as uneven coating of the specimen. .

It should be noted that a calculation method other than the above-mentioned Histogram Intersection can be used for calculating the similarity. As another example, the Bhattacharyya distance shown in the following formula (2) can also be used. In the following equation (2), “p ₁ ” indicates an appearance probability obtained by dividing the frequency of the class i of the frequency distribution H _{1 to be} evaluated by the total frequency, and “H ₂ ” is the class of the reference frequency distribution H ₂ . It refers to the probability of occurrence when the frequency of i is divided by the total frequency. Also by this Bhattacharyya distance, the determination part 15f can calculate the similarity degree which takes the value of 0 or more and 1 or less similarly to said Histogram Intersection.

  The output unit 15g is a processing unit that executes output control of information to an external device such as the client terminal 30 or the head office terminal 50.

  As one embodiment, the output unit 15g can output the colony count result to the client terminal 30 when the colony count is completed, or when the client terminal 30 or the head office terminal 50 receives a request, Colony count results can also be output on demand. Needless to say, any form of information output to the client terminal 30 or the head office terminal 50, for example, display output, audio output, or print output may be employed.

  As another embodiment, the output unit 15g outputs the inspection quality determination result by the determination unit 15f. For example, the output unit 15g outputs the test result “OK” to the client terminal 30 or the head office terminal 50 when the determination unit 15f determines that the similarity is greater than or equal to the threshold value. At this time, when the output destination of the information is the head office terminal 50, the output unit 15g refers to the above-described implementation data and obtains information on the inspector who performed the sanitary inspection of the petri dish or the group to which the inspector belongs. It can also be output. In addition to this, it is also possible to add a message regarding the risk that an accident or deficiency is low. When the output destination of the information is the client terminal 30, the output unit 15g can add a message regarding business such as shipment OK. On the other hand, the output unit 15g outputs the test result “NG” to the client terminal 30 or the head office terminal 50 when the determination unit 15f determines that the similarity is less than the threshold value. At this time, when the output destination of the information is the head office terminal 50, the output unit 15g refers to the above-described implementation data and obtains information on the inspector who performed the sanitary inspection of the petri dish or the group to which the inspector belongs. It can also be output. In addition to this, a message related to an alert that there is a high possibility that an accident or deficiency has occurred can be added. Thereby, for example, interviews can be carried out with the person in charge or the person in charge of the inspection department at the base 3. For this reason, it is possible to investigate the cause of the problem in the sanitary inspection, to take countermeasures, and to instruct the responsible person or person in charge of the inspection department that is causing the deterioration of the inspection quality. When the output destination of the information is the client terminal 30, the output unit 15g A. (Laboratory Accident) A message for instructing the creation of a log, a message for instructing a re-examination, or a message regarding work such as shipping NG can be further added. Thereby, the person in charge in the inspection department at the base 3 can self-control the inspection quality. For example, if the person in charge of the inspection department A. It is possible to create a log, input comments on how to deal with inspection quality abnormalities in reports such as history and monthly reports, perform re-inspection, and promptly collect food for shipment.

  The control unit 15 can be realized by causing a CPU (Central Processing Unit), an MPU (Micro Processing Unit), or the like to execute a colony counting program. Moreover, said control part 15 is realizable also by hard wired logics, such as ASIC (Application Specific Integrated Circuit) and FPGA (Field Programmable Gate Array).

[Process flow]
Next, the process flow of the server device 10 according to the present embodiment will be described. Here, after (1) colony counting processing executed by the server device 10 is described, (2) determination processing is described.

(1) Colony Counting Process FIG. 6 is a flowchart showing the procedure of the colony counting process according to the first embodiment. This process is activated when a petri dish image is received from the client terminal 30. The flowchart shown in FIG. 6 illustrates the processing when one petri dish image is received. However, when a plurality of petri images are received, the processing shown in the figure is executed in order or the processing shown in parallel is performed. Can be executed.

  As illustrated in FIG. 6, when the receiving unit 15a receives a petri dish image (step S101), the detection unit 15b refers to the luminance value of each pixel included in the petri dish received in step S101, thereby displaying a histogram of luminance values. Is created (step S102).

  Subsequently, the detection unit 15b calculates the luminance value having the highest frequency in the histogram of luminance values obtained in step S102 as the luminance value representing the background portion such as the culture medium, and the luminance value of the pixel in the petri dish image and A threshold value to be compared with the luminance difference between the representative luminance values of the background portion is calculated. In this way, the representative luminance value and threshold value of the background portion are set as colony detection criteria (step S103).

  Then, the detecting unit 15b generates a binarized image from the petri dish according to the colony detection criterion calculated in step S103 (step S104). Subsequently, the detection unit 15b detects colonies included in the binarized image from the binarized image generated in step S104 (step S105).

  Thereafter, the counting unit 15c counts the colonies detected in step S105 (step S106), and stores the petri image, the count result of the colonies, and the data associated with the position of each colony in the inspection data 13a in the storage unit 13 Are additionally registered (step S107), and the process is terminated. The petri dish image, the colony count result, and the position of each colony are output to the client terminal 30 that uploaded the petri dish image.

(2) Determination Processing FIG. 7 is a flowchart illustrating a determination processing procedure according to the first embodiment. As an example, this process is started when an inspection quality browsing request is received from the head office terminal 50. In addition, when the counting result by the counting unit 15c is obtained, the processing can be started, and the processing is started for the inspection data 13a whose inspection quality is not determined at a certain period or regular time. You can also.

  As illustrated in FIG. 7, the dividing unit 15 d reads the inspection data corresponding to the inspection number that has received the inspection quality browsing request from the head office terminal 50 among the inspection data 13 a stored in the storage unit 13 (Step S <b> 301). In addition, the dividing unit 15d divides the petri dish included in the previously read inspection data into a plurality of regions (step S302).

  Thereafter, the calculation unit 15e refers to the position of each colony included in the inspection data, and counts the number of colonies present in the region for each region divided by the dividing unit 15d, thereby making the evaluation quality evaluation object. The frequency distribution of the number of colonies in the petri image is calculated (step S303). Subsequently, the calculation unit 15e further derives a frequency distribution of the reference colony number to be compared with the frequency distribution of the colony number to be evaluated for the inspection quality (step S304).

  Then, the determination unit 15f calculates the similarity between the frequency distribution of the evaluation target calculated in step S303 and the reference frequency distribution derived in step S304 (step S305).

  Here, when it is determined that the similarity is equal to or greater than the threshold (Yes in Step S306), the output unit 15g outputs the inspection result “OK” to the client terminal 30 or the head office terminal 50 (Step S307), and performs the processing. finish.

  On the other hand, when it is determined that the similarity is less than the threshold (No in step S306), the output unit 15g outputs the inspection result “NG” to the client terminal 30 or the head office terminal 50 (step S308), and ends the process. To do.

[Effect of Example 1]
As described above, the server apparatus 10 according to the present embodiment estimates the distribution status of colonies on the culture surface of the petri dish from the position of the colonies on the petri image, determines the quality of the hygiene inspection based on the distribution status, and Output the judgment result. Therefore, according to the server device 10 according to the present embodiment, the inspection quality can be evaluated.

  Although the embodiments related to the disclosed apparatus have been described above, the present invention may be implemented in various different forms other than the above-described embodiments. Therefore, another embodiment included in the present invention will be described below.

[Application example of standard frequency distribution]
In the first embodiment, the case where the total frequency of the frequency distribution to be evaluated is evenly allocated to each region is illustrated, but it does not necessarily have to be equal. For example, when the division value obtained by dividing the total frequency by the number of areas has a value after the decimal point, for the frequency in the area within the predetermined range from the edge of the petri dish, the value after the decimal point of the division value is cut and the center of the petri dish As for the frequency in the region within the predetermined range from the above, the value after the decimal point of the division value may be rounded up to an integer. This reduces the distribution of the edge of the petri dish, which is difficult to apply the specimen to the petri dish medium, and distributes the central part of the petri dish, which is easier to apply the specimen than the edge of the petri dish, to other areas. Can be weighted a lot. For this reason, it is possible to use a reference frequency distribution close to the actual situation at the inspection site for calculating the similarity. As an example of a method for obtaining such a reference frequency distribution, the reference frequency distribution may be statistically calculated from the distribution of past normally cultured image data.

[Application example 1 of division]
In the first embodiment, the case where the petri dish image is divided into a lattice shape is illustrated, but the shape of dividing the petri dish image is not limited to this. For example, the dividing unit 15d can also divide the petri dish into a plurality of regions defined by the distance from the center. FIG. 8 is a diagram showing an application example of the dividing method, FIG. 9 is a diagram showing an example of the frequency distribution of the evaluation target, and FIG. 10 is a diagram showing the columnar graph of the frequency distribution shown in FIG. is there.

  As shown in FIG. 8, the dividing unit 15 d divides into five regions by five concentric circles having different distances from the center. Among these, the shape of the first region is defined by the smallest circle having the smallest diameter among the five concentric circles, and the region from the center to the distance d1 is the category. In the second region, a range obtained by removing the smallest circle from the second circle is defined by the smallest circle and the second circle having the second smallest diameter among the five concentric circles. In other words, the area from the distance d1 to the distance d2 is the category of the second area. In the third region, a range obtained by removing the second circle from the third circle is defined by the third circle and the second circle of the five concentric circles. In other words, the third region is a category from the distance d2 to the distance d3. Similarly, the area from the distance d3 to the distance d4 falls within the category of the fourth area. Further, the fifth region similarly falls within the range from the distance d4 to the distance d5.

  When such division is made, the frequency distribution of the number of colonies shown in FIG. 8 is as shown in FIG. That is, as shown in FIG. 9, four colonies are distributed in the first region where the distance d from the center is “0 ≦ d <1,” and the distance d from the center is “1 ≦ d. Eleven colonies are distributed in the second region of <2. Furthermore, in the third region where the distance d from the center is “2 ≦ d <3”, 14 colonies are distributed, and four distances d from the center are “3 ≦ d <4”. Nineteen colonies are distributed in the eye region, and 26 colonies are distributed in the fifth region where the distance d from the center is “4 ≦ d <5”. This is illustrated in a columnar graph as shown in FIG. As shown in FIG. 10, since the area becomes larger as the distance d from the center becomes larger, unlike the division method shown in the first embodiment, the stepwise columnar shape becomes closer to the uniform colony distribution. As a result, a graph is formed. Even when such division is performed, the determination unit 15f can calculate the similarity in the same manner using the above formulas (1) and (2). Thus, when the petri dish image is divided into a plurality of areas defined by the distance from the center, the petri dish portion can be included without leaving each area.

[Application example 2 of division]
For example, the dividing unit 15d can also divide the petri dish into a fan shape. FIG. 11 is a diagram illustrating an application example of the division method, and FIG. 12 is a diagram illustrating an example of the frequency distribution of the evaluation target. For example, the dividing unit 15d uses the right direction of the horizontal line as a base line (0 °), and a circle that forms a petri dish is divided at a predetermined angle, 45 ° in the illustrated example. When dividing into a fan shape in this way, as shown in FIG. 12, since the areas are equally divided between the regions, the colony distribution is uniform as in the dividing method shown in the first embodiment. As a result, a columnar graph having substantially the same size is formed. Thus, also when dividing | segmenting a petri dish image in fan shape, a petri dish part can be included without remaining in each area | region.

[Application example of threshold]
The determination unit 15f illustrated in FIG. 2 can also set a threshold to be compared with the similarity as follows. For example, the determination unit 15f can set the value based on the size of the colony count result. For example, a smaller value can be set as the threshold value as the number of colonies included in the petri dish is smaller, while a larger value can be set as the threshold value as the number of colonies is larger. This is because the degree of similarity tends to be small when the number of colonies is small or when the number of divided areas is large.

  Here, as an example, when the possibility of erroneous detection is approached to zero, a threshold value can be set with the following policy. FIG. 13 is a graph showing an example of the relationship between the number of colonies and the degree of similarity. FIG. 13 shows a graph in which the lowest similarity (Histogram Intersection) value is plotted after performing 1000 trials under a uniform distribution for each number of colonies. From the graph shown in FIG. 13, when the number of colonies is 0 to 25, the alert of the test result “NG” is not output, and when the number of colonies is 26 to 50, the similarity is 0.5. When the number of colonies is 51 to 100, an alert is output when the degree of similarity is less than 0.65, and when the number of colonies is 101 to 200 The threshold is set so that an alert is output when the similarity is less than 0.75, an alert is output when the number of colonies is 201 or more, and the similarity is less than 0.85. Can be set variable. Thus, erroneous detection can be suppressed by setting the minimum similarity included in the section of the number of colonies as a threshold value. Although the case where the threshold is determined by a uniform distribution simulation using random numbers is illustrated here, it can also be determined by statistics from actual data. For example, with respect to an image that can be regarded as being normally cultured, as shown in FIG. 13, the number of colonies and the result of histogram intersection can be plotted, and a threshold value can be determined from the tendency.

In addition, when a certain erroneous detection is allowed, the threshold value can be set by the following policy. In this case, as an example, the threshold value can be calculated according to the following equation (3). In this equation (3), “N” indicates the total frequency of the frequency distribution H of similarity, and “E _allowed ” indicates an allowable rate of false detection. FIG. 14 is a diagram illustrating an example of a frequency distribution of similarities. FIG. 14 shows a frequency distribution H of similarity (Histogram Intersection) when the number of colonies is “100”, the number of divisions of the region is “10”, and 10,000 trials are performed. For example, when setting a threshold value that allows erroneous detection up to 10%, x is obtained according to the following equation (3). As a result, it is found that the degree of similarity up to the filled portion shown in FIG. 14 corresponds to less than 10% of the total frequency, and the similarity of the portion after filling corresponds to 10% or more of the total frequency. In this case, the degree of similarity up to the filled portion shown in FIG. 14, that is, the degree of similarity “0.83” of x obtained as a solution can be set as the threshold value. In the example of FIG. 14, the threshold setting method when the number of colonies is 100 is illustrated. However, the threshold is set individually by performing the same test as the test shown in FIG. 14 for each number of colonies. preferable.

[Distribution and integration]
In addition, each component of each illustrated apparatus does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. Can be integrated and configured. For example, the receiving unit 15a, the detecting unit 15b, the counting unit 15c, the dividing unit 15d, the calculating unit 15e, the determining unit 15f, or the output unit 15g may be connected as an external device of the server device 10 via a network. In addition, each device has a receiving unit 15a, a detecting unit 15b, a counting unit 15c, a dividing unit 15d, a calculating unit 15e, a determining unit 15f, or an output unit 15g. You may make it implement | achieve the function of the server apparatus 10. FIG.

[Judgment program]
The various processes described in the above embodiments can be realized by executing a prepared program on a computer such as a personal computer or a workstation. In the following, an example of a computer that executes a determination program having the same function as that of the above embodiment will be described with reference to FIG.

  FIG. 15 is a schematic diagram illustrating an example of a computer that executes a determination program according to the first and second embodiments. As illustrated in FIG. 15, the computer 100 includes an operation unit 110 a, a speaker 110 b, a camera 110 c, a display 120, and a communication unit 130. Further, the computer 100 includes a CPU 150, a ROM 160, an HDD 170, and a RAM 180. These units 110 to 180 are connected via a bus 140.

  As shown in FIG. 15, a determination program 170a that exhibits the same functions as those of the detection unit 15b, the division unit 15d, the calculation unit 15e, the determination unit 15f, and the output unit 15g described in the first embodiment is stored in the HDD 170 in advance. Remembered. The determination program 170a may be integrated or separated as appropriate, like the respective components of the detection unit 15b, the division unit 15d, the calculation unit 15e, the determination unit 15f, and the output unit 15g illustrated in FIG. In other words, all data stored in the HDD 170 need not always be stored in the HDD 170, and only data necessary for processing may be stored in the HDD 170.

  Then, the CPU 150 reads the determination program 170 a from the HDD 170 and expands it in the RAM 180. Thereby, as shown in FIG. 15, the determination program 170a functions as a determination process 180a. The determination process 180a expands various data read from the HDD 170 in an area allocated to itself on the RAM 180 as appropriate, and executes various processes based on the expanded data. The determination process 180a includes processing executed by each functional unit shown in FIG. 2, for example, processing shown in FIG. 6 and FIG. In addition, each processing unit virtually realized on the CPU 150 does not always require that all processing units operate on the CPU 150, and only a processing unit necessary for the processing needs to be virtually realized.

  Note that the determination program 170a is not necessarily stored in the HDD 170 or the ROM 160 from the beginning. For example, each program is stored in a “portable physical medium” such as a flexible disk inserted into the computer 100, so-called FD, CD-ROM, DVD disk, magneto-optical disk, or IC card. Then, the computer 100 may acquire and execute each program from these portable physical media. In addition, each program is stored in another computer or server device connected to the computer 100 via a public line, the Internet, a LAN, a WAN, etc., and the computer 100 acquires and executes each program from these. It may be.

DESCRIPTION OF SYMBOLS 1 Inspection support service 10 Server apparatus 11 Communication I / F part 13 Memory | storage part 13a Inspection data 15 Control part 15a Reception part 15b Detection part 15c Counting part 15d Division | segmentation part 15e Calculation part 15f Determination part 15g Output part 30A, 30B, 30C Client terminal 50 Head office terminal

Claims (6)

By processing the captured image of the petri dish in which the bacteria are cultured, the position of the bacterial colonies cultured in the petri dish is detected for each of the plurality of colonies,
Based on the detected position of each of the plurality of colonies, to calculate the surface distribution of the plurality of colonies,
It is determined whether the calculated distribution status satisfies a predetermined distribution condition,
Outputting the result of the determination,
A determination method executed by a computer.   The determination method according to claim 1, wherein the determination process determines that the predetermined distribution condition is not satisfied when the distribution state can be regarded as not uniform.   3. The output process according to claim 1, wherein the output process outputs information indicating that the captured image is not suitable for processing when the result of the determination does not satisfy a condition. The determination method described. Furthermore, the captured image is stored in association with information identifying a person in charge or a group in charge of performing the bacterial culture work on the petri dish,
As a result of the determination, specify a person in charge or a group in charge corresponding to the petri dish determined to not satisfy the distribution condition,
4. The determination method according to claim 1, wherein the specified information is output. By processing the captured image of the petri dish in which the bacteria are cultured, the position of the bacterial colonies cultured in the petri dish is detected for each of the plurality of colonies,
Based on the detected position of each of the plurality of colonies, to calculate the surface distribution of the plurality of colonies,
It is determined whether the calculated distribution status satisfies a predetermined distribution condition,
Outputting the result of the determination,
A judgment program that causes a computer to execute processing. By detecting the captured image of the petri dish in which the bacteria are cultured, and detecting the position of the bacterial colonies cultured in the petri dish for a plurality of colonies, and
Based on the detected position of each of the plurality of colonies, a calculation unit that calculates a surface distribution state of the plurality of colonies;
A determination unit that determines whether the calculated distribution condition satisfies a predetermined distribution condition;
And an output unit that outputs the result of the determination.

Images