JP6229365B2 - Colony counting device, colony counting method, and colony counting program - Google Patents

Description

  This case relates to a colony counting device, a colony counting method, and a colony counting program.

  Since the bacteria are small and difficult to see directly, colonies are generated by culturing the bacteria. Patent Document 1 discloses a technique for drawing a normal line to a tangent line at each point on a contour line and counting colonies from a location where the intersections are concentrated.

Japanese Patent Laid-Open No. 9-140395

  However, in Patent Document 1, the number of normals to be drawn increases, and processing time is required.

  In one aspect, an object of the present invention is to provide a colony counting device, a colony counting method, and a colony counting program that can easily count colonies.

In one aspect, the colony counting device includes an image acquisition unit that acquires an image of cultured bacteria, a peak detection unit that detects a luminance peak in the bacteria image, and a luminance peak detected by the peak detection unit. A counting unit that counts colonies based on the number , wherein the counting unit calculates a luminance gradient around each luminance peak from a luminance distribution around the plurality of luminance peaks, and the luminance value having the lower luminance value in the luminance gradient. When a straight line is drawn from the luminance value to the higher luminance value, it is determined whether or not the change in luminance value exceeds the predetermined threshold from each luminance peak to the intersection of the straight line. In this case, the plurality of luminance peaks are counted as one luminance peak .

  Colonies can be counted easily.

(A) And (b) is a figure for demonstrating a colony. (A) is a figure for demonstrating a lecithinase reaction, (b) is a figure for demonstrating glossiness. (A) is a block diagram for demonstrating the hardware constitutions of the colony count apparatus which concerns on Example 1, (b) is a block diagram of each function implement | achieved by execution of a colony count program. It is a figure for demonstrating an example of the flowchart performed by a colony count apparatus. (A)-(c) is a figure showing the detail of a detection of a luminance peak. It is a figure showing the range of Formula (2) and Formula (3). It is an example of the flowchart performed when detecting a luminance peak. It is a figure showing the example in which several glossiness appears in one colony. It is a block diagram of each function implement | achieved by execution of the colony count program which concerns on Example 2. FIG. It is a figure for demonstrating an example of the flowchart performed by the colony count apparatus which concerns on Example 2. FIG. (A) And (b) is a figure for demonstrating a center direction. (A)-(e) is a figure for demonstrating luminance distribution. It is an example of the flowchart performed when detecting whether a brightness | luminance peak exists in the same colony.

  Prior to the description of the examples, an outline of the bacteria test will be described. At restaurants and food manufacturers, the amount of bacteria is inspected. Since the bacteria are small and cannot be seen directly, referring to FIG. 1 (a), for example, the food to be examined is diluted and applied onto the medium and cultured for several days. As a result, referring to FIG. 1B, bacteria grow and colonies having a visible size are generated. By counting colonies, the amount of bacteria can be examined. However, the method of counting colonies visually is inefficient. Therefore, a technique for counting colonies more easily is desired.

  For example, a method of automatically detecting colonies using an apparatus can be mentioned. A technique can be used in which an image of the cultured bacteria is acquired and the connected colonies are separated using the contour shape. This technique is a method of obtaining the number of the centers when the perpendiculars of the contour lines are extended and they intersect. However, this technique has a problem that it cannot be separated well when three or more colonies are connected.

  In particular, bacteria such as Staphylococcus aureus cause a phenomenon called lecithinase reaction in which the periphery of the colony becomes translucent and cloudy when cultured. FIG. 2A is a schematic diagram for explaining the lecithinase reaction. Referring to FIG. 2A, when a lecithinase reaction occurs, a translucent white turbid region 2 appears around the colony 1. When the lecithinase reaction proceeds and the cloudy region 2 spreads, the cloudy region 2 overlaps between the plurality of colonies 1. In this case, it is difficult to count colonies based on the contour line.

  On the other hand, referring to FIG. 2B, a pearly luster 3 appears in the colony 1 in a lipase-positive bacterium such as Staphylococcus aureus or an S (Smooth) type bacterium having a smooth surface. . Therefore, colonies can be counted by detecting gloss 3 and counting gloss 3. In the following examples, a colony counting device, a colony counting method, and a colony counting program capable of simply counting colonies will be described with a focus on gloss.

  FIG. 3A is a block diagram for explaining a hardware configuration of the colony counting device 100 according to the first embodiment. With reference to Fig.3 (a), the colony count apparatus 100 is provided with CPU101, RAM102, the memory | storage device 103, the display apparatus 104, the image sensor 105, the light source 106 grade | etc.,.

  A CPU (Central Processing Unit) 101 is a central processing unit. The CPU 101 includes one or more cores. A RAM (Random Access Memory) 102 is a volatile memory that temporarily stores programs executed by the CPU 101, data processed by the CPU 101, and the like. The storage device 103 is a nonvolatile storage device. As the storage device 103, for example, a ROM (Read Only Memory), a solid state drive (SSD) such as a flash memory, a hard disk driven by a hard disk drive, or the like can be used. The storage device 103 stores a colony count program and the like.

  The display device 104 is a liquid crystal display, an electroluminescence panel, or the like, and displays an image acquired by the image sensor 105, a colony count result, and the like. The image sensor 105 is not particularly limited as long as it can acquire an image of cultured bacteria, and is, for example, a CMOS (Complementary Metal Oxide Semiconductor) camera. The light source 106 is a device that irradiates light to cultured bacteria.

  As an example, the image sensor 105 acquires an image after performing culture processing on a petri dish with a medium such as an egg yolk-added mannitol salt agar medium. The position of the petri dish is predetermined. For example, there is a method in which a circle of the analysis range is superimposed and displayed on the image being photographed so that a person can perform rough positioning. In addition, a method of analyzing an image and extracting a petri dish circle using an existing method such as Hough transform is also conceivable.

  The colony count program stored in the storage device 103 is developed in the RAM 102 so as to be executable. The CPU 101 executes a colony count program developed in the RAM 102. Thereby, the colony counting process by the colony counting device 100 is executed. The colony counting process is a process of counting colonies.

  FIG. 3B is a block diagram of each function realized by executing the colony count program. By executing the colony count program, the colony count device 100 implements the image acquisition unit 10, the peak point detection unit 20, and the bacteria detection unit 30.

  Subsequently, the operation of the colony counting apparatus 100 will be described. FIG. 4 is a diagram for explaining an example of a flowchart executed by the colony counting device 100. With reference to FIG. 4, the image acquisition part 10 acquires the image of a microbe from the image sensor 105 (step S1). Next, the peak point detection unit 20 detects gloss by detecting a luminance peak as a gloss candidate from the image acquired in step S1 (step S2). Next, the bacteria detection unit 30 counts colonies by counting the gloss detected in step S2 (step S3). Colonies can be counted by executing the above processing.

  Next, details of gloss detection will be described. Gloss is a place where the brightness is higher than the surrounding area. Therefore, in this embodiment, the gloss is detected by detecting the luminance peak. FIG. 5A to FIG. 5C are diagrams showing details of detection of a luminance peak. With reference to Fig.5 (a), the peak point detection part 20 acquires the distribution of the brightness | luminance in the said image by scanning the image acquired from the image sensor 105. FIG. FIG. 5B is an enlarged view of a region including a certain colony 1. FIG.5 (c) is an enlarged view of the periphery of the glossy 3 of the colony 1 of FIG.5 (b). Referring to FIG. 5C, since the brightness of gloss 3 is higher than that of the surroundings, gloss 3 can be detected by detecting a brightness peak. For example, when a predetermined pixel (which may be one pixel or may be an average of a plurality of pixels such as 3 × 3) is higher than a threshold value compared with the surrounding pixels, the predetermined pixel The portion can be detected as a luminance peak.

First, in the image acquired by the image sensor 105, when the luminance of the pixel at the coordinate position (x, y) is p (x, y), the luminance P (x, y) is set as the luminance of the central pixel. As the luminance P (x, y) of the central pixel, the luminance p (x, y) itself may be used, or an average of a plurality of pixels may be used. For example, when calculating the average of one pixel in the vertical and horizontal directions (9 pixels in total), the luminance P (x, y) of the central pixel can be calculated according to the following formula (1).

上記式では、四角形の範囲を抽出しているが、他の範囲でも構わない。例えば円形の場合は、式は以下のようになる。
ｐ（ａ，ｂ）≦Ｐ（ｘ，ｙ） ａ，ｂは（ａ−ｘ）^２＋（ｂ−ｙ）^２≦ｒ１^２を満たす範囲。
ｐ（ｃ，ｄ）≦Ｐ（ｘ，ｙ）−ｔ ｃ，ｄは（ｒ１^２≦（ｘ−ｃ）^２＋（ｙ−ｄ）^２≦ｒ２^２を満たす範囲。 Next, the pixel values at positions where the brightness of the surrounding pixels (for example, pixels in the 10 × 10 area, excluding the vicinity of the center) are all lower than the brightness P (x, y) of the center pixel and are more than a certain distance apart. Is lower than the luminance of the central pixel by a threshold value or more, that portion is detected as a luminance peak. Specifically, when the following formula (2) and the following formula (3) are satisfied, the center pixel is detected as a luminance peak. Note that “r1” and “r2” represent ranges, and the relationship r1 <r2 is established. “T” represents a threshold value. FIG. 6 is a diagram illustrating ranges of the following formula (2) and the following formula (3).

In the above formula, a square range is extracted, but other ranges may be used. For example, in the case of a circle, the equation is as follows.
p (a, b) ≦ P (x, y) a, b satisfies ^{(a-x) 2 + (} b-y) 2 ≦ r1 2 range.
p (c, d) ≦ P (x, y) −t c, d is a range satisfying (r1 ² ≦ (x−c) ² + (y−d) ² ≦ r2 ² .

  FIG. 7 is an example of a flowchart executed when a luminance peak is detected. Referring to FIG. 7, peak point detection unit 20 sets the current position (x, y) (step S11). Next, the peak point detection unit 20 obtains an average value of pixel values around the current position (step S12). Next, the peak point detection unit 20 determines whether or not there are pixels in the vicinity of the current position that are equal to or larger than the average value obtained in step S12 (step S13).

  If “No” is determined in step S13, the peak point detection unit 20 is a pixel located at a certain distance from the current position, and whether there is a pixel equal to or greater than (average value obtained in step S12−threshold value). Determination is made (step S14). When it is determined as “No” in step S14, the peak point detection unit 20 sets the current position (x, y) as a luminance peak (step S15). Next, the peak point detector 20 moves the current position in the x-axis plus direction (eg, right) (step S16). If the current position is already at the right end, the current position is moved to the left end of the y-axis negative row (for example, down). Next, the peak point detection unit 20 determines whether or not the luminance peak detection process has been completed for all the coordinates of the image (step S17). If “No” is determined in step S17, the process is executed again from step S12.

  By executing the flowchart of FIG. 5, the luminance peak can be detected for all coordinates. If “Yes” is determined in step S13 or step S14, step S17 is executed.

  According to the present embodiment, colonies can be counted simply by detecting a luminance peak in the acquired image, so that colonies can be counted easily. In addition, even if a lecithinase reaction occurs, since the gloss appears in each colony, the colonies can be counted with high accuracy.

  A pearly luster appears in cocci such as Staphylococcus aureus, but a plurality of lusters 3 may appear in one colony 1 with reference to FIG. Therefore, in the second embodiment, colonies where a plurality of glosses appear are detected, and the number of overlapping glosses is subtracted to improve colony counting accuracy.

  FIG. 9 is a block diagram of each function realized by executing the colony count program according to the second embodiment. By executing the colony count program, the colony count device 100 includes an image acquisition unit 10, a peak point detection unit 20, a bacteria detection unit 30, a proximity peak point detection unit 40, a peak point direction detection unit 50, and a peak point direction intersection detection unit. 60, the point-to-point change detection unit 70, and the same colony determination unit 80 are realized.

  FIG. 10 is a diagram for explaining an example of a flowchart executed by the colony counting device 100 according to the second embodiment. Referring to FIG. 10, the image acquisition unit 10 acquires a fungus image from the image sensor 105 (step S <b> 21). Next, the peak point detection unit 20 detects gloss by detecting a luminance peak as a gloss candidate from the image acquired in step S21 (step S22). Next, the same colony determination unit 80 determines that the luminance peak has the same colony according to the processing results of the proximity peak point detection unit 40, the peak point direction detection unit 50, the peak point direction intersection detection unit 60, and the inter-point change detection unit 70. It is determined whether or not (step S23). Next, the bacteria detection unit 30 counts colonies by subtracting the overlap detected in step S23 from the number of luminance peaks detected in step S22 (step S24). By executing the above processing, colonies can be counted with high accuracy.

  Next, details of the process in step S23 of FIG. 10 will be described. In step S23, when a plurality of luminance peaks are detected, it is determined whether or not they are the same colony. The determination of whether or not they are the same colony can be performed by, for example, the distance and direction between two luminance peaks, and pixels.

(1) Detection of the center direction of a colony First, the proximity peak point detection unit 40 detects a plurality of luminance peaks whose distance is equal to or less than a threshold value. The peak point direction detection unit 50 detects the central direction of the colony as seen from these luminance peaks. Since the brightness around the colony is lower than the brightness within the colony, the center direction can be detected by obtaining the higher direction from the lower brightness around the gloss. That is, the center direction can be detected by obtaining the luminance gradient. FIG. 11A is a diagram illustrating a luminance gradient in each gloss 3. The following method can be considered as a method of detecting the luminance gradient. First, referring to FIG. 11B, the pixel information around the luminance peak is acquired, and the average of the luminance is detected for the areas of the designated sizes on the upper side, the lower side, the right side, and the left side. If each of them is Au, Ad, Ar, and Al, the luminance gradient direction (gradient angle θ) can be expressed by the following equation (4).

ここで、ｒ１は、中心から平均を求める内側の線までの距離、ｒ２は、中心から平均を求める外側までの距離となる。領域のサイズが不変である場合は分母を省略しても構わない。 Ar can be calculated, for example, according to the following formula (5).

Here, r1 is the distance from the center to the inner line for obtaining the average, and r2 is the distance from the center to the outer side for obtaining the average. If the size of the region is unchanged, the denominator may be omitted.

同一コロニー判定部８０は、この交点と（ｘ１，ｙ１）、（ｘ２，ｙ２）との距離を求め、その距離があらかじめ指定したしきい値以下の場合、この２つの輝度のピークが同一コロニー上にある可能性があると判定する。すなわち、同一コロニー判定部８０は、複数の輝度ピークが所定のしきい値以下となるように近接している場合に、それらの輝度ピークが同一のコロニー内のものであると判定する。なお、図１２（ａ）の例では、中心方向が交差する交点と輝度ピークとの距離が大きくなるため、検出された複数の輝度ピークは同一のコロニー内のものであると判定されない。 (2) Crossing determination of lines connecting luminance peaks The peak point direction crossing detection unit 60 obtains a point where the central directions cross each other. This intersecting point can be obtained by the intersection of straight lines in the central direction. When the positions of the luminance peaks are (x1, y1) and (x2, y2) and the gradients are a1 and a2, respectively, the intersection of the two vertices is obtained by the following formula (6) and the following formula (7). Can do.

The same colony determination unit 80 obtains the distance between this intersection and (x1, y1), (x2, y2), and when the distance is equal to or less than a predetermined threshold, the two luminance peaks are on the same colony. It is determined that there is a possibility. That is, the same colony determination unit 80 determines that the luminance peaks are within the same colony when the plurality of luminance peaks are close to each other so as to be equal to or less than a predetermined threshold value. In the example of FIG. 12A, since the distance between the intersection at which the center direction intersects and the luminance peak increases, the detected plurality of luminance peaks are not determined to be within the same colony.

(3) Extraction of luminance change on a straight line For example, when a plurality of luminance peaks appear in the same colony as shown in FIG. 12B, from one luminance peak to the intersection in the center direction as shown in FIG. And the change in luminance from the intersection to the other luminance peak becomes small. On the other hand, when the luminance peak appears in different colonies as shown in FIG. 12C, the luminance change from one luminance peak to the intersection in the central direction as shown in FIG. The change in the luminance to the luminance peak increases in the vicinity of the colony boundary. Therefore, if there is no portion that exceeds a predetermined threshold in the change in luminance from one luminance peak to the intersection in the central direction and the change in luminance from the intersection to the other luminance peak, those luminance peaks are the same. It can be determined that it appears in the colony.

First, the point-to-point change detection unit 70 calculates a straight line bent at the intersection (x, y) in (x1, y1) → (x, y) → (x2, y2). A luminance change is obtained when the coordinate position is expressed by the following formula (8) and the following formula (9). Note that t is a parameter having a value in the range of 0 ≦ t ≦ 1, and the following equation (8) is a coordinate value when t changes from 0 to 0.5 in the range of 0 ≦ t ≦ 0.5. Is an expression for obtaining the coordinate value (x ′, y ′) when fluctuating from (x1, y1) to (x, y), and the following expression (9) is in the range of 0.5 ≦ t ≦ 1. This is an equation for obtaining a coordinate value (x ′, y ′) when the coordinate value changes from (x1, y1) to (x, y) when t changes from 0.5 to 1.

In this case, the function of the luminance of the line connecting the luminance peaks can be expressed as the following formula (10). In the following formula (10), t is 0 ≦ t ≦ 1. Further, v (t) is a luminance value at the position t. In this case, the variance of luminance can be obtained according to the following formula (11). The v bar is an average value of v (t). Although the equation (11) uses an integral, the variance may be calculated as a sum of discrete values as in the equation (12). Here, m is the number of points to be calculated, and a fixed value such as 10 is designated. When the value of this variance is smaller than a predetermined threshold value, it can be determined that they are on the same colony.

  It may be possible to determine whether or not they are in the same colony by using a simplified method depending on conditions. For example, it is conceivable that the determination is made up to the processing of (2) without using the luminance change condition of (3). It is also conceivable to use the luminance change (3) on a straight line connecting two luminance peaks without using the processes (1) and (2).

  FIG. 13 is an example of a flowchart executed when detecting whether or not the luminance peak is in the same colony. Referring to FIG. 13, the proximity peak point detection unit 40 selects two luminance peaks that are close in distance (below the threshold value) (step S31). Next, the peak point direction detection unit 50 obtains the luminance average of the upper, lower, left and right pixels of each luminance peak, and obtains the luminance gradient (center direction) (step S32).

  Next, the peak point direction intersection detection unit 60 obtains the intersections of the two luminance gradients, and obtains the respective distances (step S33). The point-to-point change detection unit 70 determines whether or not the intersection is at least a threshold with respect to the two luminance peaks (step S34). When it is determined as “No” in step S34, the point-to-point change detection unit 70 obtains a luminance change on the intersection line (step S35).

  Next, the same colony determination unit 80 determines whether the luminance change is equal to or less than a threshold value (step S36). When it determines with "Yes" by step S36, the same colony determination part 80 determines with two brightness | luminance peaks being in the same colony (step S37). When it is determined as “No” in step S36, the same colony determination unit 80 determines that the two luminance peaks are not in the same colony (step S38). If it is determined “Yes” in step S34, step S38 is executed. After execution of step S37 or step S38, the same colony determination unit 80 determines whether all luminance peaks have been determined (step S39). If it is determined “No” in step S39, step S31 is executed again. If “Yes” is determined in step S39, the execution of the flowchart ends.

  According to the present embodiment, when a plurality of glosses appear in the same colony, the overlap can be excluded from the count. Thereby, colony counting accuracy is improved. Specifically, when a straight line is drawn in the center direction from two luminance peaks whose distance is less than or equal to the threshold value, if the distance between the intersection of the straight line and the two luminance peaks is equal to or less than the threshold value, It is determined that the two luminance peaks are in the same colony. In this case, the colony counting accuracy is improved. Further, in a straight line connecting the two luminance peaks and the intersection, when the luminance change is large, it is determined that the two luminance peaks are in different colonies, and when the luminance change is small, the two luminance peaks Are determined to be in the same colony. Also in this case, the colony counting accuracy is improved.

  In addition, each part implement | achieved by execution of a colony count program may be comprised by the circuit for exclusive use.

  In Example 1, the bacteria detection unit 30 functions as a counting unit that counts colonies. In the second embodiment, the proximity peak point detection unit 40, the peak point direction detection unit 50, the peak point direction intersection detection unit 60, the point-to-point change detection unit 70, the same colony determination unit 80, and the bacteria detection unit 30 count colonies. Functions as a counting unit.

  Although the embodiments of the present invention have been described in detail above, the present invention is not limited to such specific embodiments, and various modifications and changes can be made within the scope of the gist of the present invention described in the claims. It can be changed.

DESCRIPTION OF SYMBOLS 10 Image acquisition part 20 Peak point detection part 30 Bacteria detection part 40 Proximity peak point detection part 50 Peak point direction detection part 60 Peak point direction crossing detection part 70 Point change detection part 80 Same colony determination part 100 Colony count apparatus

Claims (5)

An image acquisition unit for acquiring an image of the cultured bacteria;
In the image of the fungus, a peak detection unit that detects a luminance peak;
A counting unit that counts colonies based on the number of luminance peaks detected by the peak detection unit , and
The counting unit calculates a luminance gradient around each luminance peak from a luminance distribution around a plurality of luminance peaks, and draws a straight line from a lower luminance value to a higher luminance value in the luminance gradient. It is determined whether or not the change in luminance value exceeds a predetermined threshold value from the peak to the intersection of the straight line, and when it is determined that it does not extend, the plurality of luminance peaks are regarded as one luminance peak. colonies count and wherein the counting. The colony counting device according to claim 1, wherein a distance between the plurality of luminance peaks is equal to or less than a predetermined threshold value. The colony counting device according to claim 1 or 2, wherein the bacterium is Staphylococcus aureus. Obtain an image of the cultured bacteria,
In the image of the fungus, a luminance peak is detected,
Count colonies based on the number of detected luminance peaks ,
At the time of counting, the brightness gradient around each brightness peak is calculated from the brightness distribution around the plurality of brightness peaks, and when a straight line is drawn from the lower brightness value to the higher brightness value in the brightness gradient, It is determined whether or not the change of the luminance value exceeds a predetermined threshold value from the luminance peak to the intersection of the straight line, and when it is determined that it does not extend, the plurality of luminance peaks are regarded as one luminance peak. Counting as a colony counting method. On the computer,
Obtaining an image of the cultured bacteria;
Detecting a luminance peak in the bacteria image;
Counting colonies based on the number of detected luminance peaks , and
In the counting step, a luminance gradient around each luminance peak is calculated from a luminance distribution around a plurality of luminance peaks, and each line is drawn when a straight line is drawn from a lower luminance value to a higher luminance value in the luminance gradient. It is determined whether or not the change of the luminance value exceeds a predetermined threshold value from the luminance peak to the intersection of the straight line, and when it is determined that it does not extend, the plurality of luminance peaks are regarded as one luminance peak. colonies count program characterized by counting as.

Images