Information processing device and its contamination detection method, water treatment system, and contamination detection program

Abstract

To reduce an influence of noise on contamination detection.SOLUTION: An information processing device (1) includes: an image acquisition unit (101) that acquires sludge images taken of a floc by a photographing device through an inspection window provided in a flocculation tank for forming the floc; and a sludge image detection unit (103) that calculates statistics of shading values for each micro-region by using multiple sludge images divided into a plurality of micro-regions and detects sludge of the inspection window based on the calculated statistical value.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an information processing device and the like that detects dirt on a window between a liquid to be treated containing solid suspended matter and an imaging device.

Background Art

[0002] Conventionally, a technique has been used in which a chemical such as a flocculant is added to a liquid to be treated such as sewage sludge and stirred to aggregate solid suspended matter to form flocs, and the aggregated sludge, which is an aggregate of these flocs, is dehydrated to obtain dehydrated sludge. In order to stably perform sludge dehydration, it is necessary to form flocs of an appropriate size. Therefore, in the process of forming flocs, it is desirable to perform control at any time to maintain the size of the flocs at an appropriate size while grasping the state of the flocs being formed.

[0003] Among the techniques for grasping the state of flocs, there is a technique that uses an image of flocs photographed by an imaging device. In this case, if the window between the flocs and the imaging device is dirty, it is difficult to appropriately grasp the state of the flocs. Therefore, it is desirable to detect the dirt on the window and promptly remove the detected dirt.

[0004] As documents that disclose techniques for detecting dirt on a window between flocs and an imaging device, for example, the following Patent Documents 1 and 2 can be cited. The floc image recognition device described in Patent Document 1 includes image processing means that performs image processing by classifying an image photographed by an imaging device into flocs and a background for each pixel. The image processing means always determines that the number of pixels indicating flocs from the start of image processing is the same as the number of those pixels in the previous image, and when this coincidence occurs twice continuously, it determines that the window is dirty. Further, the dirt detection device for an imaging device described in Patent Document 2 extracts the difference in density values of a plurality of images photographed by the imaging device, integrates the extracted differences in density values, and determines that the dirt exists in a region where the integrated difference in density values is less than or equal to a predetermined value.

Prior Art Documents

Patent Documents

[0005] [Patent Document 1] Japanese Patent Application Publication No. 62-85843 [Patent Document 2] Japanese Patent Publication No. 2003-259358 [Overview of the project] [Problems that the invention aims to solve]

[0006] The devices described in Patent Documents 1 and 2 are susceptible to noise-induced interference in stain detection. For example, in the case of Patent Document 1, if a pixel corresponding to window stain is misidentified as part of the background due to noise, it becomes difficult to detect the stain. Similarly, in the case of Patent Document 2, if the difference in grayscale values ​​in a stained area becomes large due to noise, it becomes difficult to detect the stain.

[0007] One aspect of the present invention aims to reduce the influence of noise on the detection of contamination. [Means for solving the problem]

[0008] To solve the above problems, an information processing device according to one aspect of the present invention includes an acquisition unit that acquires an image of the flocs taken by a photographing device through a window provided in a tank for forming flocs, and a detection unit that uses a plurality of images divided into a plurality of minute regions to calculate a statistical value of the density for each minute region and detects dirt on the window based on the statistical value.

[0009] Furthermore, a control method for an information processing apparatus according to another aspect of the present invention includes an acquisition step of acquiring an image of the flocs captured by a photographing device through a window provided in a tank for forming flocs, and a detection step of calculating a statistical value of the grayscale for each of the multiple images divided into multiple microregions, and detecting dirt on the window based on the statistical value. [Effects of the Invention]

[0010] According to one aspect of the present invention, the influence of noise on the detection of contamination can be reduced. [Brief explanation of the drawing]

[0011] [Figure 1] This is a block diagram showing an example of the main components of an information processing device according to one embodiment of the present invention. [Figure 2] This figure shows an example configuration of a water treatment system including the above-mentioned information processing device. [Figure 3] This figure shows an example of a sludge image that has been binarized. [Figure 4] This figure shows examples of sludge images containing flocs of different sizes that have been binarized. [Figure 5] This flowchart shows an example of the index value calculation process performed by the above-mentioned information processing device. [Figure 6] This figure shows examples of sludge images taken every few days. [Figure 7] This figure shows an overview of the multiple micro-regions into which the sludge image has been divided. [Figure 8] This graph shows the average value of the grayscale levels in each micro-region, calculated using multiple sludge images. [Figure 9] This graph shows the average value of the grayscale in each micro-region, calculated using 10 sludge images, for each sludge image shown in Figure 6. [Figure 10] This flowchart shows an example of the dirt detection process performed by the above-mentioned information processing device. [Figure 11] This is a block diagram showing an example of the main components of an information processing device according to another embodiment of the present invention. [Figure 12] This flowchart shows an example of the dirt detection process performed by the above-mentioned information processing device. [Modes for carrying out the invention]

[0012] Hereinafter, embodiments of the present invention will be described in detail. For the sake of convenience of explanation, members having the same functions as those shown in each embodiment are given the same reference numerals, and the description thereof will be omitted as appropriate.

[0013] [Embodiment 1] An embodiment of the present invention will be described with reference to FIGS. 1 to 10.

[0014] (Water treatment system) The outline of the water treatment system 100 according to this embodiment will be described based on FIG. 2. FIG. 2 is a diagram showing a configuration example of the water treatment system 100. The water treatment system 100 is a system for separating a liquid to be treated into solid suspensions and a liquid component. As shown in the figure, it includes an information processing device 1, a control device 3, a flocculator 5, an addition device 6, and a dehydrator 9. Hereinafter, an example in which the liquid to be treated is sludge generated by biological treatment such as sewage will be described. Sludge is a liquid containing solid suspensions and can also be called slurry.

[0015] The flocculator 5 is a device that aggregates solid suspensions in sludge to form flocs and obtains aggregated sludge. The flocculator 5 in FIG. 2 includes an aggregation tank (also called a stirring tank) 51 (tank), a stirring blade 52, a motor 53, and an inspection window 54 (window). Since the sludge in the aggregation tank 51 is stirred by rotating the stirring blade 52 by the motor 53, the flocculator 5 can also be said to be a stirring device. In addition, the flocculator 5 is provided with a sludge inlet 55, a chemical inlet 56, and an outlet 57.

[0016] Furthermore, an imaging device 72 and an imaging lighting device 71 are attached to the inspection window 54. The imaging device 72 only needs to be capable of capturing still images. It is preferable that the coagulation tank 51 be opaque so that the way light hits the flocs does not change while the water treatment system 100 is in operation. It is also preferable that the imaging device 72 and the lighting device 71 are housed in a light-shielding dark box with an opening on the inspection window 54 side, as shown in the illustrated example, and that the remaining part of the inspection window 54 is also shielded from light, at least during imaging. The imaging device 72 may be used to take images from above the liquid surface. Alternatively, the imaging device 72 may be submerged in water for imaging.

[0017] The dewatering machine 9 is installed downstream of the flocculator 5 and is a device that dewaters the flocculated sludge discharged from the flocculator 5. The dewatering machine 9 in Figure 2 is a screw press type dewatering machine equipped with an outer shell screen 91 and a screw 92. The dewatering machine 9 is also provided with a sludge inlet 93, a filtrate outlet 94, and a dewatered cake outlet 95. Of course, the dewatering machine 9 can be any machine capable of dewatering flocculated sludge and is not limited to a screw press type. For example, a centrifugal dewatering machine, a filter press type dewatering machine, or a belt press dewatering machine can also be used.

[0018] In the water treatment system 100, the sludge to be treated is continuously or intermittently supplied from the sludge inlet 55 into the flocculator 5's coagulation tank 51 by a supply device (not shown). The sludge supply rate may be automatically controlled by the supply device or its control device according to the sludge processing rate of the flocculator 5 and the dewatering machine 9.

[0019] Then, the additive device 6, based on the control of the control device 3, introduces a chemical agent (at least containing a coagulant) for coagulating sludge into the coagulation tank 51 from the chemical inlet 56. In this state, the motor 53 is driven to rotate the stirring blade 52, stirring the sludge and chemical agent in the coagulation tank 51 and forming flocs. The coagulated sludge, which is a mixture of the formed flocs and the water contained in the sludge, is then discharged from the discharge port 57. The chemical agent may be added in advance to the sludge before it is introduced into the flocculator 5 from the sludge inlet 55.

[0020] During the process of forming these flocs, the imaging device 72 captures images. Since these images contain sludge, they will be referred to as sludge images below. Multiple flocs overlap in the sludge images, making it difficult to calculate the number of flocs or the area of ​​each individual floc. Therefore, the information processing device 1 detects the background area rather than the flocs from the sludge image and analyzes it to calculate an index value indicating the state of floc formation.

[0021] Next, the coagulated sludge is supplied into the outer shell screen 91 from the sludge inlet 93 of the dewatering machine 9. Inside the dewatering machine 9, the coagulated sludge is dewatered under pressure from the screw 92, the filtrate is discharged from the filtrate outlet 94, and the dewatered cake, which is a mass of dewatered coagulated sludge, is discharged from the dewatered cake outlet 95.

[0022] The control device 3 controls at least one of the following: the amount of chemical added by the additive device 6 and the stirring speed in the flocculator 5, according to the index value indicating the floc formation state calculated by the information processing device 1, so that the size of the flocs becomes appropriate. Details of this control will be described later. The control device 3 may also control other equipment in the water treatment system 100. The stirring speed can also be referred to as the stirring intensity.

[0023] As described above, the water treatment system 100 includes an additive device 6 for adding an agent to the sludge in the coagulation tank 51 that coagulates solid suspended matter, a flocculator 5 for stirring the sludge in the coagulation tank 51, an imaging device 72 for photographing the flocs in the coagulation tank 51 through an inspection window 54, an information processing device 1 for calculating an index value indicating the floc formation state from the image captured by the imaging device 72, and a control device 3 for controlling at least one of the amount of agent added by the additive device 6 and the stirring speed according to the calculated index value.

[0024] As described above, the information processing device 1 detects the background area rather than the flocs themselves from the sludge image and calculates an index value indicating the floc formation state by analyzing it. Therefore, it is possible to calculate an index value that accurately indicates the floc formation state from a sludge image in which multiple flocs overlap. Furthermore, controlling the amount of chemical added and controlling the stirring speed are both effective in changing the size of the flocs. Thus, the water treatment system 100 can automatically improve the floc formation state while treating the sludge and stably generate flocs of an appropriate size.

[0025] (Information processing device) A more detailed description of the information processing device 1 will be given based on Figure 1. Figure 1 is a block diagram showing an example of the main components of the information processing device 1. As shown in Figure 1, the information processing device 1 includes a control unit 10 that controls all parts of the information processing device 1, and a storage unit 11 that stores various data used by the information processing device 1. The information processing device 1 also includes a communication unit 12 for the information processing device 1 to communicate with other devices (e.g., a control device 3), an input unit 13 that receives input to the information processing device 1, and an output unit 14 for the information processing device 1 to output information.

[0026] The control unit 10 includes an image acquisition unit 101 (acquisition unit), an image analysis unit 102 (processing unit), a dirt detection unit 103 (detection unit), and an instruction unit 104. The storage unit 11 stores the sludge image 111 and the binarized image 112, and also includes an index value storage unit 113. Furthermore, the storage unit 11 contains micro-region information 114. The dirt detection unit 103 and the micro-region information 114 will be described later.

[0027] The image acquisition unit 101 successively acquires sludge images 111 from the imaging device 72 via the communication unit 12. The image acquisition unit 101 stores the acquired sludge images 111 in the storage unit 11. Therefore, multiple sludge images 111 are stored in the storage unit 11.

[0028] The image analysis unit 102 (processing unit) performs image processing to detect the floc formation state for each sludge image 111 stored in the storage unit 11. The image analysis unit 102 includes a binarization processing unit 1021, a background detection unit 1022, and an index value calculation unit 1023.

[0029] The binarization processing unit 1021 binarizes the sludge image 111 stored in the storage unit 11 and generates a binarized image 112 divided into a flock region and a background region. The binarization processing unit 1021 then stores the generated binarized image 112 in the storage unit 11. The sludge image 111 and the binarized image 112 will be described later based on Figure 3.

[0030] The background detection unit 1022 detects the background region of a flock from an image in which multiple flocks overlap. Specifically, the background detection unit 1022 acquires the binarized image 112 stored in the storage unit 11 and detects the background region in the binarized image 112. The background detection unit 1022 outputs the background region detection result to the index value calculation unit 1023. As will be described in detail later, the background region is composed of multiple sub-regions.

[0031] The index value calculation unit 1023 analyzes multiple sub-regions constituting the background region detected by the background detection unit 1022 and calculates an index value indicating the floc formation state. The index value calculation unit 1023 also stores the calculated index value in the index value storage unit 113 of the storage unit 11. The calculation of the index value will be described later.

[0032] The instruction unit 104 issues various instructions to the control device 3 via the communication unit 12. Based on these instructions, the control device 3 controls various devices. Alternatively, the instruction unit 104 may control the devices via the communication unit 12. In this case, the control device 3 can be omitted.

[0033] In this embodiment, the instruction unit 104 instructs the control device 3 to maintain the floc size appropriately based on the index value stored in the index value storage unit 113. The control device 3 controls, for example, at least one of the amount of chemical added by the additive device 6 in the water treatment system 100 and the rotation speed of the motor 53 based on the instruction from the instruction unit 104.

[0034] (Regarding sludge images and binarized images) Figures 3 and 4 show sludge images A1 to A5, each containing flocs of different sizes, and examples of binarized images a1 to a5 obtained by binarizing each sludge image. Sludge image A1 shown in Figure 3 is generally a dark image, with the floc areas appearing whiter than their background areas.

[0035] In the binarization process of the sludge image A1, the binarization processing unit 1021 determines whether the pixel value of each pixel constituting the sludge image A1 is equal to or greater than a predetermined threshold. The binarization processing unit 1021 then sets pixels in the sludge image A1 that are determined to be equal to or greater than the threshold to white, and pixels that are determined to be less than the threshold to black.

[0036] This generates the binarized image a1 shown in Figure 3. Binarized image a1 is an image in which the flock region 301 is represented in white and its background region 300 is represented in black. In this way, in binarized image a1, the boundary between the flock region and the background region, which was unclear in the sludge image A1, can be clearly recognized. For example, from binarized image a1, the gaps between flocks and the steps in the depth direction are clearly recognized as the background region 300, and the outlines and irregularities of the flocks can also be clearly recognized as the background region 300.

[0037] The sludge images A1 to A5 shown in Figures 3 and 4 are classified by visual inspection as "extremely large," "large," "medium," "small," and "extremely small," respectively. As shown in Figures 3 and 4, as the floc size approaches "extremely small" from "extremely large," the background region 300 becomes subdivided; that is, the area of ​​each of the multiple small regions constituting the background region 300 becomes smaller and the number of such regions increases. Therefore, the index value calculation unit 1023 can calculate an index value indicating the floc formation state by analyzing the multiple small regions constituting the background region.

[0038] For example, the index value calculation unit 1023 may calculate the average area of ​​the sub-regions as the index value. In this case, the index value calculation unit 1023 first counts the number of sub-regions by considering a continuous region consisting of adjacent pixels among the pixels included in the background region detected by the background detection unit 1022 as one sub-region. Then, the index value calculation unit 1023 calculates the index value as the value obtained by dividing the total area of ​​the background region by the number of sub-regions, i.e., the average area of ​​the sub-regions.

[0039] The number of small regions in the binarized images a1 to a5 shown in Figures 3 and 4 is 36, 49, 57, 71, and 69, respectively, and the average area, or index value, is 980, 933, 546, 392, and 231 pixels, respectively. Thus, the average area decreases as the flock size decreases. Therefore, the average area of ​​the small regions is a valid index value.

[0040] The analysis of the sub-regions is not limited to the above example, as long as it yields an index value indicating the floc formation state. For example, the index value calculation unit 1023 may determine the area, major side length, minor side length, circumcircle radius or diameter, incircle radius or diameter, or perimeter for each sub-region through the analysis of the sub-regions. The index value calculation unit 1023 may then calculate the average value obtained by averaging the values ​​across all sub-regions as the index value. The major side length and minor side length are the lengths of the major and minor sides when the sub-region is approximated as an ellipse. The circumcircle radius or diameter is the radius or diameter of the smallest circumcircle surrounding the sub-region. The incircle radius or diameter is the radius or diameter of the largest incircle surrounding the sub-region.

[0041] (Regarding instructions based on indicator values) The instruction unit 104 instructs the control device 3 to maintain the flock size appropriately based on the index value calculated as described above. For example, the instruction unit 104 may instruct the control device 3 to perform control to return the index value to the appropriate range when the index value falls outside a predetermined appropriate range.

[0042] If the indicator value falls below the appropriate range, i.e., if the floc size is "extremely small," the instruction unit 104 can instruct the control device 3 to increase the size of the flocs formed. For example, if the amount of flocculant added is sufficient, generally, reducing the stirring speed of the flocculator 5 will increase the size of the flocs formed, so the instruction unit 104 may instruct the control device 3 to reduce the rotation speed of the motor 53. Also, if the amount of flocculant added is insufficient, generally, increasing the flocculant injection rate will increase the size of the flocs formed, so the instruction unit 104 may instruct the control device 3 to increase the amount of flocculant added to the additive device 6. Both of these instructions may be given simultaneously or at different times. In the latter case, for example, the instruction unit 104 may first instruct the control device 3 to control the motor 53, and then, if the indicator value does not fall within the appropriate range after a predetermined time has elapsed since that control, instruct the control device 3 to control the additive device 6.

[0043] Similarly, if the indicator value exceeds the above appropriate range, that is, if the floc size falls between "extremely large" and "medium", the instruction unit 104 can instruct the control device 3 to reduce the size of the floc formed. For example, the instruction unit 104 may instruct the control device 3 to increase the rotational speed of the motor 53, or to reduce the amount of coagulant added to the additive device 6.

[0044] Furthermore, the instruction unit 104 may instruct the control device 3 to change the controlled object or controlled amount according to the degree of deviation from the appropriate range. For example, the instruction unit 104 may instruct the control device 3 to increase the amount of control for the rotational speed of the motor 53 according to the degree of deviation from the appropriate range of the index value (difference between the index value and the upper limit of the appropriate range / difference between the lower limit of the appropriate range and the index value). Also, for example, the instruction unit 104 may instruct the control device 3 to control both the motor 53 and the additive device 6 when the floc size is "extremely large" or "large," and to control only the motor 53 when it is "medium" or "extremely small."

[0045] The appropriate size of the flocs may vary depending on the type of sludge and the type of dewatering machine. Therefore, the appropriate range mentioned above should be predetermined according to the type of sludge, the type of dewatering machine, etc. Furthermore, the control for adjusting the size of the flocs is not limited to the examples given above, and the instruction unit 104 may instruct the control device 3 to perform any control that affects the size of the flocs for any equipment that affects the size of the flocs.

[0046] Furthermore, if automatic control is not required, the instruction unit 104 may omit giving instructions to the control device 3 according to the index value calculated by the index value calculation unit 1023. In this case, the instruction unit 104 may make the user aware of the index value by outputting it to the output unit 14, and may instruct the control device 3 to perform manual control to adjust the flock size based on the input from the user via the input unit 13. In this case, the index value calculation unit 1023 may classify the size of the flock using the calculated index value. The index value calculation unit 1023 may also output a classification such as "extremely large" to the output unit 14 along with a numerical value such as the average area of ​​the small region, or instead of a numerical value. This makes it easy for the user to recognize the size of the flock.

[0047] (Calculation process for indicator values) The flow of the index value calculation process (index value calculation method) performed by the information processing device 1 will be explained based on Figure 5. Figure 5 is a flowchart of an example of the index value calculation process performed by the information processing device 1. The processes S1 to S6 described below are performed continuously while the water treatment system 100 is in operation. In addition, while the water treatment system 100 is in operation, the imaging device 72 takes pictures at predetermined intervals, and the captured images are stored as sludge images 111 in the storage unit 11 of the information processing device 1.

[0048] In S1, the binarization processing unit 1021 acquires the sludge image 111 from the storage unit 11. If multiple sludge images 111 are stored, the binarization processing unit 1021 acquires the latest sludge image 111. Then, in S2, the binarization processing unit 1021 binarizes the sludge image 111 acquired in S1 to generate a binarized image, and stores this as the binarized image 112 in the storage unit 11.

[0049] In S3, the background detection unit 1022 acquires the binarized image 112 that was generated in S2 and stored in the storage unit 11. As described above, the binarized image 112 contains multiple overlapping flocks (see, for example, Figures 3 and 4). The background detection unit 1022 then detects the background region of the flocks from the acquired binarized image 112.

[0050] In S4, the index value calculation unit 1023 analyzes the multiple sub-regions that make up the background region detected in S3 and calculates an index value indicating the floc formation state. For example, the index value calculation unit 1023 may count the number of multiple sub-regions that make up the background region, calculate the total area of ​​the background region, and then calculate the index value as the value obtained by dividing the calculated total area by the number of sub-regions, i.e., the average area of ​​the sub-regions.

[0051] In S5, the instruction unit 104 determines whether the control device 3 needs to control the equipment based on the index value calculated by the index value calculation unit 1023. Figure 5 illustrates an example where the equipment to be controlled is the motor 53, that is, an example where the floc size is adjusted by adjusting the stirring speed. Of course, the instruction unit 104 may also instruct the control device 3 to adjust the floc size by controlling the additive device 6 or other devices.

[0052] If it is determined in S5 that control is not necessary (No in S5), the process in Figure 5 ends. On the other hand, if it is determined that control is necessary (Yes in S5), the process proceeds to S6. The criteria for determining whether control is necessary can be predetermined; for example, the indicator unit 104 may determine that control is necessary if the index value is outside a predetermined appropriate range.

[0053] In S6, the instruction unit 104 instructs the control device 3 to change the number of times the floc is stirred per unit time. Specifically, the instruction unit 104 changes the number of stirs by instructing the control device 3 to change the rotation speed of the motor 53, and this completes the process shown in Figure 5. The method for changing the rotation speed has already been explained, so it will not be repeated here.

[0054] (Overview of the dirt detection unit) Next, the dirt detection unit 103 will be described based on Figures 6 to 9. In this embodiment, the dirt detection unit 103 detects permeable dirt adhering to the inspection window 54.

[0055] For example, ferric polysulfate, also known as polyiron, is a reddish-brown liquid used as a flocculant. When ferric polysulfate is used as a flocculant, permeable fouling may adhere to a portion of the inspection window 54.

[0056] The sludge image 111 described above is an image of the flocs in the coagulation tank 51, captured by the imaging device 72 through the inspection window 54. Therefore, if transparent dirt adheres to a part of the inspection window 54, the area of ​​the sludge image corresponding to that area of ​​the inspection window 54 will be darker than other areas of the sludge image, and may be mistakenly detected as the background area of ​​the flocs by the background detection unit 1022. In other words, if transparent dirt adheres to a part of the inspection window 54, the index value may not be calculated appropriately.

[0057] Figure 6 shows examples of sludge images B1 to B4 taken at intervals of several days. In the first sludge image B1 shown in Figure 6, no transparent contamination was observed by visual inspection.

[0058] In the second sludge image B2 shown in Figure 6, slight translucent contamination was visible to the naked eye. Specifically, the left central part of the second sludge image B2 was slightly darkened. However, this left central part was not mistakenly detected as a floc background area by the background detection unit 1022. Therefore, the second sludge image B2 did not affect the calculation of the index value by the index value calculation unit 1023.

[0059] In the third sludge image B3 shown in Figure 6, transparent contamination was clearly visible to the naked eye. Specifically, the central part of the third sludge image B3 was distinctly darkened. This central part was then mistakenly detected by the background detection unit 1022 as the background area of ​​the floc. Consequently, the third sludge image B3 affected the calculation of the index value by the index value calculation unit 1023.

[0060] The fourth sludge image B4 shown in Figure 6 was taken the day after the inspection window 54 was cleaned. Similar to the first sludge image B1, no permeable contamination was visible in the fourth sludge image B4. Therefore, it is desirable to be able to detect permeable contamination that would affect the calculation of the above index values.

[0061] To detect the transparent dirt mentioned above, the dirt detection unit 103 uses multiple sludge images 111 divided into multiple minute regions to calculate statistical values ​​of the intensity for each minute region, and detects dirt on the inspection window 54 based on these statistical values. With the above configuration, even if the intensity values ​​calculated for each minute region change significantly due to noise, the dirt detection unit 103 can detect dirt on the window based on the statistical values ​​of the intensity, thereby reducing the impact of noise on the detection.

[0062] (Details of the dirt detection unit) As shown in Figure 1, the stain detection unit 103 includes a grayscale conversion unit 1031, a density value calculation unit 1032, an average value calculation unit 1033, a variance value calculation unit 1034, and a stain determination unit 1035.

[0063] The grayscale conversion unit 1031 acquires the sludge image 111, which is a color image (R, G, B), from the storage unit 11 and converts it to a grayscale image (Y) using the following equation (1). Y=0.299R+0.587G+0.114B...(1). Note that a color image may be converted to a grayscale image using any formula other than formula (1) above. The grayscale conversion unit 1031 sends the grayscale image (Y) to the grayscale value calculation unit 1032.

[0064] Next, the grayscale value calculation unit 1032 calculates a grayscale value for each of the minute regions using the grayscale image (Y) from the grayscale conversion unit 1031 and the minute region information 114 from the storage unit 11. The grayscale value calculation unit 1032 sends the calculated grayscale values ​​for each minute region to the average value calculation unit 1033.

[0065] Figure 7 shows an overview of the multiple microregions into which the sludge image 111 is divided. The microregion information in the storage unit 11 contains the number and position information of each microregion. In the example in Figure 7, the sludge image 111 has a size of 640 pixels × 478 pixels.

[0066] The sludge image 111 includes an edge 1111 with a width of approximately 20 pixels. The edge 1111 is susceptible to the effects of illumination from the lighting device 71. For this reason, the edge 1111 is not used in the calculation of the above index value. Therefore, since the edge 1111 does not need to be used for detecting the above-mentioned dirt, the above-mentioned minute region is not assigned to it.

[0067] In the sludge image 111, the target area 1112, excluding the edges 1111, is divided into multiple micro-regions MA. In the example in Figure 7, the micro-regions MA are 32 pixels × 32 pixels in size and are arranged in 13 rows × 18 columns. Therefore, there are 234 micro-regions MA. The micro-regions MA are numbered from 0 to 233, starting from the top left and ending in the bottom right.

[0068] As shown in Figure 7, the lower and right edges of the target area 1112 have a width of less than 32 pixels, so it is not possible to assign a micro-region MA to them. Therefore, the lower and right edges are not used for detecting dirt. Also, if the size of the sludge image 111 is 640 pixels × 480 pixels, the size of the micro-region MA is 32 pixels × 32 pixels, and the entire area of ​​the sludge image 111 is considered the target area 1112, then the number of micro-region MAs will be 300. Furthermore, the size of the micro-region MA is not limited to 32 pixels × 32 pixels.

[0069] The grayscale value calculation unit 1032 uses the average value of the grayscale values ​​of multiple pixels included in the microregion MA as the grayscale value of the microregion MA. This allows for obtaining 234 grayscale values ​​for microregions from a single sludge image 111. The grayscale value calculation unit 1032 may also use any statistical value, such as the median, mode, maximum, or minimum of the grayscale values ​​of the multiple pixels, as the grayscale value of the microregion MA.

[0070] The average value calculation unit 1033 uses the most recent multiple sludge images 111 to obtain the density values ​​of each micro-region MA from the density value calculation unit 1032 and calculates the average value of the density values ​​for each micro-region MA. The average value calculation unit 1033 sends the calculated average value of the density values ​​of each micro-region MA to the variance value calculation unit 1034.

[0071] Figure 8 is a graph showing the average intensity values ​​in each micro-region MA, calculated using multiple recent sludge images 111. In Figure 8, graphs G1, G2, G3, and G4 are graphs using 1, 5, 10, and 20 sludge images 111, respectively. The sludge images 111 used in Figure 8 are images taken immediately after cleaning the inspection window 54 (corresponding to sludge image B4 in Figure 6), that is, images taken when the inspection window 54 was free of dirt.

[0072] Referring to Figure 8, it can be seen that as the number of sludge images 111 used increases, the variation in the average value of the grayscale values ​​is suppressed. On the other hand, as the number of sludge images 111 used increases, the processing time in the grayscale conversion unit 1031, the grayscale value calculation unit 1032, and the average value calculation unit 1033 increases. Therefore, the number of sludge images 111 used should preferably be 5 to 20, more preferably 8 to 15, and even more preferably 10. For example, if the imaging device 72 takes a picture once every 6 minutes, the most recent 10 sludge images 111 can be acquired in 1 hour.

[0073] Figure 9 is a graph showing the average intensity values ​​in each micro-region MA, calculated using 10 sludge images 111. Graphs G11 to G14 in Figure 9 correspond to sludge images B1 to B4 in Figure 6, respectively.

[0074] Referring to Figure 9, it can be seen that the variation in the average value of the grayscale values ​​is greater in the second graph G12 than in the first graph G11, and greater in the third graph G13 than in the second graph G12. In other words, it can be seen that the variation in the average value of the grayscale values ​​increases as the permeability of the inspection window 54 deteriorates. Referring to the fourth graph G14 in Figure 9, it can be seen that by cleaning the inspection window 54, the variation in the average value of the grayscale values ​​becomes as small as in the first graph G11 in Figure 9.

[0075] Therefore, the variance calculation unit 1034 uses the average value of the intensity values ​​of each minute region MA from the average value calculation unit 1033 to calculate the variance of the above average values ​​in multiple minute region MAs. The variance calculation unit 1034 sends the calculated variance to the stain detection unit 1035.

[0076] Furthermore, referring to Figure 9, there are minute MA regions numbered 0 and 17 where the average value of the grayscale values ​​deviates significantly compared to others. This deviation in the average value of these minute MA regions is thought to be due to causes other than the contamination of the translucency, such as the effect of illumination by the lighting device 71. For this reason, these minute MA regions are thought to have a negative impact on the calculation of the variance value.

[0077] Therefore, it is desirable for the variance calculation unit 1034 to calculate the variance of the average values ​​after excluding a predetermined number of average values ​​from the largest average value and a predetermined number of average values ​​from the smallest average value when the average values ​​of the above intensity values ​​in all minute regions MA are arranged. In this case, average values ​​that are thought to have an adverse effect on the calculation of the variance are excluded from the variance calculation process, so that the permeability of the inspection window 54 can be detected with even greater accuracy.

[0078] The above predetermined number may be selected from 4 to 6, or from 1 to 3% of the total number of micro-region MAs (234). As a result of the above calculation, the above variance values ​​for the above sludge images B1 to B4 were 47, 59, 120, and 46, respectively.

[0079] The stain detection unit 1035 determines whether or not permeable stains have been detected based on the variance of the average value from the variance value calculation unit 1034. If the stain detection unit 1035 determines that stains have been detected, it notifies the instruction unit 104 of the detection of stains.

[0080] Since the variance value corresponding to sludge image B3 in Figure 6, which affected the calculation of the above index value, is 120, the contamination detection unit 1035 should determine that it has detected permeable contamination that affects the calculation of the above index value when the above variance value is 100 or more.

[0081] Incidentally, when two types of flocs with different average sizes were mixed in the coagulation tank 51, and multiple sludge images 111 were taken by the imaging device 72 through the inspection window 54 where no permeable fouling was attached, the variance of the average value of the above-mentioned density values ​​was at most 68. Therefore, the fouling detection unit 1035 may determine that permeable fouling has been detected when the above-mentioned variance is 80 or higher.

[0082] When the dirt detection unit 103 notifies the output unit 14 of the detection of the dirt, the instruction unit 104 instructs the output unit 14 to issue an alarm. This allows the user to be alerted to the dirt. The instruction unit 104 also changes the instruction given to the control device 3 from an automatic instruction based on the index value calculated by image processing in the image analysis unit 102 to a manual instruction based on user input in the input unit 13. This prevents inappropriate instructions from being given to the control device 3 based on inappropriate index values ​​calculated due to dirt on the inspection window 54.

[0083] Furthermore, based on the above warning, the user may stop the water treatment system 100 and clean the inspection window 54. Alternatively, based on the above warning, the user may stop the information processing device 1, visually inspect the liquid to be treated in the coagulation tank 51 through the inspection window 54 to determine the state of floc formation, and then operate the control device 3 to adjust the additive device 6 and the stirring device 5, etc., based on the result of the above determination. In this case, the operation of the water treatment system 100 can be continued.

[0084] As described above, the dirt detection unit 103 uses a plurality of sludge images 111 stored in the memory unit 11 to calculate the average value of the intensity for each of the minute regions, calculates the variance of the average values ​​in the plurality of minute regions, and detects the permeability of the inspection window 54 based on the variance.

[0085] According to the above configuration, the average value of the density of each micro-region MA converges to a certain value as the number of sludge images 111 increases. At this time, in micro-region MAs containing permeable dirt, the converged value is lower than in micro-region MAs that do not contain dirt. Therefore, in inspection windows 54 where permeable dirt exists in some areas, the variance value is larger than in inspection windows 54 where no dirt exists. Also, the variance value increases as the permeability of the dirt decreases. On the other hand, when the permeability of the dirt is high, the impact on image processing for detecting the floc formation state, i.e., the calculation of the index value, is small. Therefore, it is possible to detect permeable dirt in inspection windows 54 that would affect the image processing based on the variance value.

[0086] Furthermore, the information processing device 1 performs both the detection of floc formation by the image analysis unit 102 and the detection of dirt on the inspection window 54 by the dirt detection unit 103 in a single unit. This eliminates the need to install two separate devices with image processing capabilities, resulting in a less expensive device configuration. Additionally, by separating the timing of floc formation detection and the timing of dirt detection on the inspection window 54, the processing load on the information processing device 1 can be reduced.

[0087] (Dirt detection process) Figure 10 is a flowchart showing an example of a stain detection process. The processes S11 to S17 described below can be performed at any frequency from every few seconds to every few days.

[0088] In S11, the grayscale conversion unit 1031 converts the multiple color sludge images 111 stored in the storage unit 11 into multiple grayscale sludge images 111 using the above formula (1). Note that if the multiple sludge images 111 stored in the storage unit 11 are grayscale images, S11 is omitted.

[0089] Next, in S12, the grayscale sludge value calculation unit 1032 calculates a grayscale value for each sludge image 111 and for each micro-region MA using the multiple grayscale sludge images 111 and the micro-region information 114 of the storage unit 11. Next, in S13, the average value calculation unit 1033 calculates the average value of the grayscale values ​​for each micro-region MA in the multiple sludge images 111.

[0090] Next, in S14, the variance calculation unit 1034 excludes a predetermined number of average values ​​from the maximum average value and a predetermined number of average values ​​from the minimum average value when the average values ​​of the above intensity values ​​in all minute regions MA are arranged. Next, in S15, the variance calculation unit 1034 calculates the variance of the remaining multiple average values.

[0091] Next, in S16, the dirt detection unit 1035 determines whether the above-mentioned variance value is equal to or greater than a threshold (for example, 100). If it is determined in S16 that the above-mentioned variance value is less than the threshold (No in S16), the dirt detection unit 1035 determines that it has not detected any permeable dirt in the inspection window 54 that would affect the calculation of the above-mentioned index value, and terminates the dirt detection process. On the other hand, if it is determined in S16 that the above-mentioned variance value is equal to or greater than the threshold (Yes in S16), the process proceeds to S17.

[0092] In S17, if the dirt detection unit 1035 detects the permeable dirt, the instruction unit 104 instructs the output unit 14 to issue an alarm, and also changes the instruction to the control device 3 from an automatic instruction based on the index value calculated in S4 of Figure 5 to a manual instruction based on the user input in the input unit 13. After that, the dirt detection process is terminated.

[0093] (modified version) The grayscale value calculation unit 1032 may store the grayscale values ​​of each minute region MA for the most recent 10 sludge images 111 in the storage unit 11. In this case, the grayscale conversion unit 1031 and the grayscale value calculation unit 1032 only need to operate on the sludge image 111 newly acquired by the image acquisition unit 101, and do not need to operate on all 10 sludge images 111.

[0094] Alternatively, the grayscale value calculation unit 1032 may discard the grayscale values ​​of each microregion MA for the oldest sludge image 111 among the 10 sludge images 111 in the storage unit 11, and store the grayscale values ​​of each microregion MA for the newly acquired sludge image 111 in the storage unit 11. In this case, the storage unit 11 only needs to secure resources to store the grayscale values ​​of each microregion MA for the most recent 10 sludge images 111, and does not need to increase the above resources even if the number of sludge images 111 stored in the storage unit 11 increases.

[0095] [Embodiment 2] Other embodiments of the present invention will be described with reference to Figures 11 and 12.

[0096] Figure 11 is a block diagram showing an example of the main components of the information processing device 1 in the water treatment system 100 according to this embodiment. The information processing device 1 of this embodiment differs from the information processing device 1 shown in Figure 1 in that the dirt detection unit 103 detects impermeable dirt adhering to the inspection window 54, but the other components are the same.

[0097] The intensity values ​​of the regions in the sludge image 111 containing the aforementioned opaque contaminants are approximately constant across multiple sludge images 111. Therefore, in minute regions MA containing opaque contaminants, the variance of intensity values ​​is smaller compared to minute regions MA without such contaminants. On the other hand, when there are few minute regions MA containing opaque contaminants, the impact on image processing for detecting the floc formation state, i.e., image processing for calculating index values, is small.

[0098] Therefore, in this embodiment, the dirt detection unit 103 uses multiple sludge images 111 divided into multiple micro-region MAs to calculate the variance value of the intensity for each micro-region MA, and detects impermeable dirt in the inspection window 54 based on the number of micro-region MAs whose variance value is below a threshold. This makes it possible to detect impermeable dirt in the inspection window 54 that would affect the image processing. The number of micro-region MAs can be set based on their impact on the image processing. The threshold may be 10, 5, or 0.

[0099] The dirt detection unit 103 shown in Figure 11 differs from the dirt detection unit 103 shown in Figure 1 in that it includes a variance calculation unit 1036, a dirt determination unit 1037, and a display instruction unit 1038 instead of the average value calculation unit 1033, variance calculation unit 1034, and dirt determination unit 1035, but the other configurations are the same.

[0100] The dispersion value calculation unit 1036 uses the most recent multiple sludge images 111 to obtain the intensity values ​​of each micro-region MA from the intensity value calculation unit 1032 and calculates the dispersion value of the intensity values ​​for each micro-region MA. The dispersion value calculation unit 1036 sends the calculated dispersion value of the intensity values ​​of each micro-region MA to the contamination determination unit 1037.

[0101] The dirt detection unit 1037 determines that it has detected impermeable dirt in the inspection window 54 if the number of minute region MAs whose variance value from the variance value calculation unit 1036 is below a threshold is greater than or equal to a predetermined number. If the dirt detection unit 1037 determines that it has detected dirt, it notifies the instruction unit 104 of the detection of dirt. In addition, if the dirt detection unit 1037 determines that it has detected dirt, it sends information about the minute region MAs whose variance value is below a threshold to the display instruction unit 1038.

[0102] The display instruction unit 1038 instructs the display device in the output unit 14 to display the information of the minute area MA from the dirt detection unit 1037. That is, if the display instruction unit 1038 detects impermeable dirt in the inspection window 54, it instructs the display device to display the information of the minute area MA whose dispersion value is below the threshold.

[0103] According to the above configuration, by referring to the information of the minute region MA where the above-mentioned variance value is below the threshold, the location of the impermeable dirt in the inspection window 54 can be easily identified. As a result, the impermeable dirt in the inspection window 54 can be efficiently removed.

[0104] Figure 12 is a flowchart showing an example of the dirt detection process. The processes S21 to S26 described below can be performed at any frequency from every few seconds to every few days. Note that S21 and S22 are the same as S11 and S12 shown in Figure 10, so their explanation will be omitted.

[0105] In S23, the variance calculation unit 1036 calculates the variance of the intensity values ​​for each micro-region MA in the multiple sludge images 111. Next, in S24, the contamination determination unit 1037 determines whether the number of micro-region MAs whose variance is below a threshold is greater than or equal to a predetermined number. If the number of micro-region MAs in S24 is less than the predetermined number (No in S24), the contamination determination unit 1037 determines that it did not detect any opaque contamination in the inspection window 54 that would affect the calculation of the index value, and terminates the contamination detection process. On the other hand, if the number of micro-region MAs in S24 is greater than or equal to the predetermined number (Yes in S24), the process proceeds to S25.

[0106] In S25, if the stain detection unit 1035 detects the impermeable stain, the instruction unit 104 instructs the output unit 14 to issue an alarm, and also changes the instruction to the control device 3 from an automatic instruction based on the index value calculated in S4 of Figure 5 to a manual instruction based on the user's input in the input unit 13. Next, in S26, the display instruction unit 1038 instructs the display device of the output unit 14 to display information on the minute region MA where the variance value is below the threshold. Note that S25 and S26 may be executed either first or simultaneously. After that, the stain detection process is terminated.

[0107] (modified version) If the number of microregion MAs (microregion MAs) containing impermeable stains, i.e., microregion MAs whose variance value from the variance value calculation unit 1036 is below a threshold, is less than a predetermined number (No in S24 of Figure 12), the stain determination unit 1037 may send information on the microregion MAs to the index value calculation unit 1023 of the image analysis unit 102. Alternatively, the index value calculation unit 1023 may use the information on the microregion MAs from the stain determination unit 1037 to analyze a plurality of small regions that constitute the background region detected by the background detection unit 1022, excluding the microregion MAs, and calculate an index value indicating the floc formation state. In this case, the index value calculation unit 1023 excludes microregion MAs containing impermeable stains from the plurality of small regions to be analyzed, so the index value can be calculated with high accuracy.

[0108] [Special Notes] The equipment configuration of the water treatment system 100 described in the above embodiment is just one example, and a water treatment system with similar functions can be constructed with various equipment configurations. Furthermore, the entities that execute each process described in the above embodiment are also just examples. Each of the above processes can be appropriately assigned to each device that constitutes the water treatment system.

[0109] For example, the water treatment system 100 may include a control panel that controls the addition of a coagulant and stirring, and an image processing device that performs binarization of sludge images. In this case, the image processing device may be configured to detect the background region and calculate the index value. The control panel may be configured to update the threshold used for binarization. Thus, each process described in the above embodiment (especially each process included in the flowchart of Figure 5) may be divided and executed by multiple information processing devices.

[0110] Furthermore, the image analysis unit 102 and the dirt detection unit 103 may be provided in separate information processing devices. In this case, the timing for detecting the floc formation state and the timing for detecting dirt can be arbitrarily set without considering the processing load of the information processing device.

[0111] [Examples of implementation using software] The functions of the information processing device 1 (hereinafter referred to as "the device") are programs that cause the device to function as a computer, and these programs can be realized by programs that cause each control block of the device (especially each part included in the control unit 10) to function as a computer.

[0112] In this case, the device includes a computer having at least one control device (e.g., a processor) and at least one storage device (e.g., memory) as hardware for executing the program. By executing the program using this control device and storage device, the functions described in each of the embodiments are realized.

[0113] The above program may be recorded on one or more computer-readable recording media, not temporary ones. These recording media may or may not be provided by the above device. In the latter case, the program may be supplied to the above device via any wired or wireless transmission medium.

[0114] Furthermore, some or all of the functions of each of the above control blocks can also be realized by logic circuits. For example, an integrated circuit in which logic circuits functioning as each of the above control blocks are formed is also included in the scope of the present invention. In addition, it is also possible to realize the functions of each of the above control blocks by, for example, a quantum computer.

[0115] Furthermore, each process described in the above embodiments may be performed by AI (Artificial Intelligence). In this case, the AI ​​may operate on the control device described above, or it may operate on other devices (for example, an edge computer or a cloud server).

[0116] The present invention is not limited to the embodiments described above, and various modifications are possible within the scope of the claims. Embodiments obtained by appropriately combining the technical means disclosed in different embodiments are also included in the technical scope of the present invention. [Explanation of symbols]

[0117] 1. Information Processing Device 3. Control device 5. Flocculator (Agitator) 6 Addition device 9 Dehydrator 10 Control Unit 11 Storage section 12 Communications Department 13 Input section 14 Output section 51 Coagulation tank (tank) 52 Agitator blades 53 Motor 54 Inspection window (window) 55 Sludge inlet 56 Drug input slot 57 Outlet 71 Lighting equipment 72 Imaging device 100 Water Treatment Systems 101 Image acquisition unit (acquisition unit) 102 Image Analysis Unit (Processing Unit) 103 Dirt detection unit (detection unit) 104 Instruction section 111 Sludge images 112 Binarized Images 113 Indicator Value Storage Unit 114 Micro area information 1021 Binarization Processing Unit 1022 Background detection unit 1023 Indicator Value Calculation Unit 1031 Grayscale conversion section 1032 Tone Value Calculation Unit 1033 Average Value Calculation Unit 1034 Variance Calculation Unit 1035 Judgment section 1036 Variance Calculation Unit 1037 Judgment section 1038 Display instruction section

Claims

1. An acquisition unit that acquires an image of the flocs captured by the imaging device through a window provided in the tank for forming the flocs, An information processing apparatus comprising: a detection unit that uses multiple images divided into multiple minute regions to calculate statistical values ​​of grayscale for each minute region, and detects dirt on the window based on the statistical values.

2. The information processing apparatus according to claim 1, further comprising a processing unit for performing image processing to detect the formation state of the floc for each of the aforementioned images.

3. The information processing apparatus according to claim 2, wherein the detection unit calculates the average value of the grayscale values ​​for each of the micro-regions using a plurality of images, calculates the variance of the average values ​​in the plurality of micro-regions, and detects the transparency of the window based on the variance.

4. The information processing apparatus according to claim 3, wherein, when calculating the variance value, the detection unit excludes a predetermined number of average values ​​from the maximum average value and a predetermined number of average values ​​from the minimum average value among the average values ​​of the intensity values ​​in the plurality of minute regions.

5. The information processing apparatus according to claim 2, wherein the detection unit calculates a variance value of the intensity values ​​for each minute region using a plurality of images, and detects the opaque dirt of the window based on the number of minute regions where the variance value is below a threshold.

6. The information processing apparatus according to claim 5, wherein the detection unit further comprises a display instruction unit that, when it detects an opaque stain in the window, instructs a display device to display information of a minute area in which the dispersion value is below a threshold.

7. An additive device for adding a chemical agent that coagulates solid suspended matter to the liquid to be treated in the tank, A stirring device for stirring the liquid to be treated in the tank, A photographic device for photographing the flocs in the tank through a window provided in the tank, An information processing device according to any one of claims 2 to 6, which detects the formation state of the floc and the dirt on the window from an image captured by the aforementioned imaging device, A water treatment system comprising: a control device that controls at least one of the additive device and the stirring device based on instructions from the information processing device corresponding to the floc formation state and the fouling of the window.

8. The water treatment system according to claim 7, wherein when the information processing device detects dirt on the window, the information processing device issues an alarm and changes the instruction to the control device from an instruction corresponding to the floc formation state detected by image processing of the image to an instruction corresponding to user input.

9. An acquisition step in which an imaging device acquires an image of the flocs through a window provided in the tank for forming the flocs, A method for detecting dirt on an information processing device, comprising: a detection step of using multiple images divided into multiple minute regions to calculate statistical values ​​of grayscale for each minute region, and detecting dirt on the window based on the statistical values.

10. A dirt detection program for causing a computer to function as an information processing device according to claim 1, wherein the dirt detection program causes the computer to function as the detection unit.

Images