Computer program, operation support apparatus, operation support system, and operation support method

Abstract

To provide a computer program capable of continuously obtaining sludge states on operation of a sludge dryer without using a water content measuring instrument, an operation support apparatus, an operation support system, and an operation support method.SOLUTION: A computer program causes a computer to run processing including: acquiring time-series images obtained by imaging the inside of a sludge dryer to which a sludge is transported while agitating, specifying a segmentation image that is obtained by segmentation of each acquired time-series image to a plurality of regions, determining whether or not there is a sludge in each specified segmentation image, and calculating sludge quantity based on the presence / absence determination result of sludge.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a computer program, a driving assistance device, a driving assistance system, and a driving assistance method. [Background technology]

[0002] Sewage sludge is often dehydrated to a specified moisture content (e.g., 80%) or less using a dehydrator such as a centrifuge to produce dehydrated sludge (also called dehydrated cake), which is then disposed of in landfills. However, as the capacity of landfill sites is decreasing year by year, dehydrated sludge is fed into a sludge dryer to produce dry sludge within the required range (e.g., 40-50%), and then fed into an incinerator for incineration.

[0003] Patent Document 1 discloses a sludge dryer that includes a cylindrical bowl and a screw blade and in which the transport speed and rolling speed of the sludge can be set. [Prior art documents] [Patent documents]

[0004] JP 2020-199470 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, in a sludge dryer such as that described in Patent Document 1, the state of the sludge inside is often visually checked through a sight glass provided in the casing of the sludge dryer so that the moisture content of the dried sludge is within a required range. However, it is not realistic to always visually check while the sludge dryer is in operation. On the other hand, there are moisture content meters that can measure the moisture content of the dewatered sludge fed into the sludge dryer and the moisture content of the dried sludge discharged from the sludge dryer, but they are expensive and do not have sufficient measurement accuracy, so in practice moisture content meters are often not used.

[0006] The present invention has been made in consideration of the above circumstances, and aims to provide a computer program, an operation support device, an operation support system, and an operation support method that can continuously grasp the state of sludge during operation of a sludge dryer without using a moisture content measuring instrument. [Means for solving the problem]

[0007] The present application includes multiple means for solving the above-mentioned problems, and one example is a computer program that causes a computer to perform the following processes: acquire a time series of images taken inside a sludge dryer where sludge is stirred and transported, identify divided images divided into a plurality of areas in each of the acquired time series images, determine the presence or absence of sludge in each of the identified divided images, and calculate the amount of sludge based on the determination result of the presence or absence of sludge. Effect of the Invention

[0008] According to the present invention, the state of sludge during operation of a sludge dryer can be continuously monitored without using a moisture content measuring instrument. [Brief description of the drawings]

[0009] [Figure 1] 1 is a diagram illustrating an example of a configuration of a driving assistance system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a diagram showing an example of the internal structure of a sludge dryer. [Diagram 3] FIG. 1 is a diagram illustrating a first example of the configuration of a driving assistance device. [Figure 4] FIG. 2 is a diagram showing an example of the inside of a sludge dryer photographed by a camera. [Diagram 5] FIG. 13 is a diagram illustrating an example of processing using a learning model. [Figure 6] FIG. 13 is a diagram showing an example of determining the presence or absence of sludge in a divided image. [Figure 7] FIG. 4 is a diagram illustrating an example of a method for calculating the amount of sludge. [Figure 8] FIG. 4 is a diagram showing an example of a transition of the amount of sludge. [Figure 9] FIG. 4 is a diagram showing an example of the relationship between the amount of sludge and the water content. [Figure 10] 11A and 11B are diagrams illustrating an example of a method for tracking feature points by an image processing unit. [Figure 11] 13A and 13B are diagrams illustrating an example of determining the presence or absence of sludge in a divided image based on feature points. [Figure 12] FIG. 2 is a diagram showing an example of the inside of a sludge dryer photographed by a camera. [Figure 13] FIG. 13 is a diagram showing an example of determining the presence or absence of sludge in a divided image. [Figure 14] FIG. 4 is a diagram illustrating an example of a method for calculating the amount of sludge. [Figure 15] The image shown here was taken by a camera of a conveyor that transports sludge discharged from a sludge dryer. [Figure 16] 11 is a diagram showing an example of relationship information indicating a correspondence relationship between a color component and a moisture content. FIG. [Figure 17] 1 is a diagram showing an example of sludge falling from a conveyor that transports sludge to be fed into a sludge dryer into a distributor. FIG. [Figure 18] FIG. 13 is a diagram showing an example of a method for identifying the position at which sludge falls from a conveyor. [Figure 19] FIG. 13 is a diagram showing the relationship between the falling position of sludge and the adhesion degree or moisture content. [Figure 20] FIG. 13 is a diagram showing an example of displaying the progress of the amount of sludge inside a sludge dryer. [Figure 21] FIG. 13 is a diagram showing an example of displaying the transition of the moisture content of sludge. [Figure 22] FIG. 2 is a diagram showing an example of an operation control screen of the sludge dryer. [Diagram 23] FIG. 4 is a diagram showing a first example of a processing procedure performed by a driving assistance device. [Figure 24] FIG. 11 is a diagram showing a second example of a processing procedure performed by the driving assistance device. [Diagram 25] FIG. 11 is a diagram illustrating a third example of a processing procedure performed by the driving assistance device. [Figure 26] FIG. 2 is a diagram illustrating a second example of the configuration of a driving assistance device. [Figure 27] The image shown here was taken by a camera of a conveyor that transports sludge discharged from a sludge dryer. [Figure 28] FIG. 13 is a diagram illustrating an example of processing using a learning model. [Figure 29] FIG. 13 is a diagram showing an example of a classified image. [Diagram 30] FIG. 13 is a diagram illustrating an example of a correspondence relationship between classes and scales. [Diagram 31] FIG. 11 is a diagram showing an example of an image of sludge at each scale. [Diagram 32] FIG. 2 is a diagram showing an example of a state of sludge. [Diagram 33] FIG. 4 is a diagram showing an example of a correspondence relationship between scale and the moisture content of sludge. [Diagram 34] FIG. 11 is a diagram showing a fourth example of a processing procedure performed by the driving assistance device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Fig. 1 is a diagram showing an example of the configuration of an operation support system of this embodiment. The operation support system includes an operation support device 50 and supports the operation of a sludge dryer 30 in a sludge treatment system. The sludge treatment system includes a dehydrator 10, a sludge dryer 30, an incinerator 20, conveyors 11, 12, and 13, a distributor 15 for inputting sludge, and the like.

[0011] The dehydrator 10 includes a centrifuge and the like. When collected sewage sludge is input, the dehydrator 10 dehydrates the input sewage sludge so that its moisture content becomes a predetermined value (e.g., 80%) or less, and transports the dehydrated sludge (also called dehydrated sludge) by a conveyor 11 attached to the outlet of the dehydrator 10. The sludge that falls from the conveyor 11 is transported by the next conveyor 12. In the conveyor 12, the sludge is weighed as necessary, and the sludge transported by the conveyor 12 falls from the tip of the conveyor 12 and is supplied to the distributor 15. A camera 21 is installed on the tip side of the conveyor 12 as a second imaging device. The camera 21 photographs the conveyor 12 that transports the sludge to be input to the sludge dryer 30. This allows the camera 21 to photograph the sludge on the conveyor 12 falling from the tip of the conveyor 12.

[0012] The distributor 15 is movable in the horizontal direction and adjusts the position at which the sludge is fed into the sludge dryer 30 depending on the moisture content of the sludge. For example, when the moisture content is high, the sludge is fed into a position further upstream (away from the discharge outlet 32) from the inlet 31 of the sludge dryer 30, and when the moisture content is low, the sludge is fed into a position further downstream (closer to the discharge outlet 32) from the inlet 31 of the sludge dryer 30.

[0013] The sludge dryer 30 has heat transfer blades fixed to a hollow rotating shaft inside a casing. A heat transfer medium such as steam, heat transfer oil, or high-temperature water is introduced into the hollow part of the rotating shaft. As a result, sludge (dewatered sludge) fed from inlet 31 is heated and dried by the rotating shaft and heat transfer blades, and is also agitated and transported by the heat transfer blades, and discharged from outlet 32. Furthermore, cameras 22, 23 as first imaging devices are attached to predetermined locations of the casing of sludge dryer 30, and cameras 22, 23 can photograph the inside of sludge dryer 30. Camera 22 can be attached, for example, near the midpoint of sludge dryer 30, and camera 23 can be attached near the exit side (outlet 32 ​​side) of sludge dryer 30. Details of sludge dryer 30 will be described later.

[0014] Sludge (also called dried sludge) discharged from the discharge outlet 32 ​​of the sludge dryer 30 is supplied to the conveyor 13. The sludge is transported by the conveyor 13 and fed into the incinerator 20. The fed sludge is combusted by the incinerator 20. The combusted sludge is discharged from the discharge outlet 25. The conveyor 13 is, for example, a flight conveyor. A camera 24 is installed above the conveyor 13 as a third imaging device. The camera 24 can capture images of the conveyor 13 that transports the sludge discharged from the sludge dryer 30.

[0015] The sludge dryer controller 40 controls the operation of the sludge dryer 30 based on the operation control signal output by the operation support device 50. The operation control of the sludge dryer 30 includes operation parameters such as the sludge processing amount (e.g., the weight of dewatered sludge fed into the sludge dryer 30 per unit time), the amount of water sprayed (e.g., the amount of water sprayed to give lubricity to the sludge received in the sludge receiving hopper and facilitate sludge transport), steam pressure (the pressure of steam as a heat medium supplied to the sludge dryer 30), the gate opening, and the flow rate of the heat medium (flow rate of steam). Note that the operation parameters are not limited to these.

[0016] The driving assistance system of the present embodiment may include a sludge dryer 30 and cameras 21, 22, 23, and 24 in addition to the driving assistance device 50.

[0017] FIG. 2 is a diagram showing an example of the internal structure of a sludge dryer 30. FIG. 2A shows the front of the sludge dryer 30, and for convenience, a part of the casing is cut away so that the inside can be shown. FIG. 2B shows a vertical cross section of the sludge dryer 30. Inside the sludge dryer 30, two rotating shafts 33 are arranged horizontally at an appropriate distance apart. The number of rotating shafts 33 is not limited to two, and may be one or three or more. The rotating shaft 33 is hollow inside and communicates with a heat medium pipe 35. The heat medium (e.g., steam) introduced from the heat medium pipe 35 flows through the hollow part of the rotating shaft 33 and is discharged from a heat medium outlet (not shown).

[0018] A plurality of heat transfer vanes (also called paddles) 34 are fixed to each rotating shaft 33 at appropriate intervals. The heat transfer vanes 34 of one rotating shaft 33 are arranged so that they fit between adjacent heat transfer vanes 34 of the other rotating shaft 33. The heat transfer vanes 34 rotate due to the rotation of the rotating shaft 33, and the sludge between the heat transfer vanes 34 is transported from the upstream side (the inlet 31 side) to the downstream side (the outlet 32 ​​side) while being stirred. As shown in FIG. 2B, the camera 23 is attached to the casing near the outlet 32 ​​and photographs the sludge inside the sludge dryer 30. Although not shown, the camera 22 is attached to the casing near the midpoint between the inlet 31 and the outlet 32 ​​and photographs the sludge inside the sludge dryer 30.

[0019] 3 is a diagram showing a first example of the configuration of the driving support device 50. The driving support device 50 includes a control unit 51 that controls the entire device, a communication unit 52, a memory 53, an image processing unit 54, an interface unit 55, a display unit 56, and a storage unit 57.

[0020] The control unit 51 may be configured by incorporating a required number of central processing units (CPUs), micro-processing units (MPUs), graphics processing units (GPUs), etc. The control unit 51 may also be configured by combining digital signal processors (DSPs), field-programmable gate arrays (FPGAs), etc.

[0021] The communication unit 52 includes a communication module and has a function of wireless communication, wired communication, or both. The communication unit 52 can be connected to the cameras 21-24 and acquires time-series images captured by the cameras 21-24. The cameras 21-24 can output time-series images at a frame rate of, for example, 60 fps, but the frame rate is not limited to 60 fps. The time-series images acquired via the communication unit 52 can be temporarily stored in the memory 53.

[0022] The image processing unit 54 performs required image processing on the time-series images acquired via the communication unit 52. The required image processing includes, for example, processing for extracting feature points indicative of the characteristics of sludge, matching processing for matching feature points in images between frames, tracking processing for tracking the matched feature points, sludge detection processing, and pixel value calculation processing for calculating color components (e.g., RGB components, etc.) and luminance values ​​of sludge.

[0023] The interface unit 55 has an interface function with the sludge dryer controller 40, and outputs a control signal for controlling (adjusting) the operation parameters of the sludge dryer 30 under the control of the control unit 51. In addition, the interface unit 55 acquires the operation parameters of the sludge dryer 30 during operation from the sludge dryer controller 40.

[0024] The display unit 56 can be configured with a liquid crystal panel, an organic EL (Electro Luminescence) display, etc. Instead of the display unit 56, or together with the display unit 56, a display device may be provided outside the driving assistance device 50.

[0025] The storage unit 57 can be configured, for example, with a hard disk or a semiconductor memory, and stores a computer program 58 (program product), a learning model 59, and required information. The learning model 59 will be described in detail later.

[0026] The computer program 58 is a computer program that runs on the driving assistance device 50, and may be downloaded from an external device via the communication unit 52 and stored in the storage unit 57. Also, the computer program 58 recorded on a recording medium (for example, an optically readable disk storage medium such as a CD-ROM) may be read by a recording medium reading unit and stored in the storage unit 57. The computer program 58 may be deployed so as to be executed on a single computer or on multiple computers interconnected by a communication network.

[0027] Memory 53 can be configured with a semiconductor memory such as a static random access memory (SRAM), a dynamic random access memory (DRAM), or a flash memory. Computer program 58 can be loaded into memory 53, and control unit 51 can execute computer program 58. Control unit 51 can execute processing defined by computer program 58. In other words, processing by control unit 51 is also processing by computer program 58.

[0028] Next, a method for continuously grasping the state of the sludge (e.g., the amount of sludge, the moisture content of the sludge, etc.) by photographing the inside of the sludge dryer 30 with the cameras 22, 23 will be described. In this embodiment, the state of the sludge is grasped by photographing the inside of the sludge dryer 30 near the discharge outlet 32, and the state of the sludge is grasped by photographing the inside of the sludge dryer 30 near the midpoint between the inlet 31 and the discharge outlet 32. First, a case of grasping the state of the sludge by photographing the inside of the sludge dryer 30 near the discharge outlet 32 ​​will be described.

[0029] FIG. 4 is a diagram showing an example of the inside of the sludge dryer 30 photographed by the camera 23. FIG. 4 shows an image photographed of the inside near the discharge port 32 of the sludge dryer 30. In the example of FIG. 4, six heat transfer vanes 34 fixed to one rotating shaft 33 are illustrated. In reality, the heat transfer vanes 34 fixed to the other rotating shaft 33 are also photographed at the same time, but for convenience, FIG. 4 illustrates only one rotating shaft 33 and the heat transfer vanes 34 fixed to that rotating shaft 33. In the photographed image, divided images D1, D2, D3, D4, and D5, which are images of the areas between adjacent heat transfer vanes 34, are identified.

[0030] In this way, the control unit 51 specifies the image of the area between the heat transfer vanes 34 that mix and transport the sludge as a divided image. The number of divided images specified in one frame of image is not limited to five. The number of divided images may be adjusted according to the number of heat transfer vanes 34 that exist within the shooting range.

[0031] That is, the control unit 51 acquires time-series images taken inside the sludge dryer 30 where the sludge is stirred and transported, and identifies divided images D1 to D5 obtained by dividing each of the acquired time-series images into a plurality of regions.

[0032] FIG. 5 is a diagram showing an example of processing by the learning model 59. For example, the learning model 59 can use a model such as a convolutional neural network (CNN), YOLO (You Only Look Once), or SSD (single shot multibox detector), and is trained (generated) to determine whether or not sludge is present in a segmented image captured inside the sludge dryer 30 when the segmented image is input. For example, as shown in FIG. 5A, when a segmented image with a relatively large amount of sludge is input to the learning model 59, the learning model 59 determines whether or not sludge is present in the segmented image, and outputs "sludge present". On the other hand, as shown in FIG. 5B, when a segmented image with a relatively small amount of sludge (or almost no sludge) is input to the learning model 59, the learning model 59 determines whether or not sludge is present in the segmented image, and outputs "no sludge". The amount of sludge present in the segmented image at which the learning model 59 outputs "sludge present" can be determined by the teacher data (association between the segmented image and the correct answer label) used when the learning model 59 is trained (generated). For example, the learning model 59 can be trained to determine that "sludge is present" if sludge is present in 50% or more of the entire region of the divided image.

[0033] As described above, the control unit 51 inputs the specified divided image to the learning model 59 and determines whether or not sludge is present in the specified divided image. The determination of the presence or absence of sludge in the divided image can be performed, for example, once per second, but the determination cycle for the presence or absence of sludge in the divided image is not limited to one second.

[0034] Fig. 6 is a diagram showing an example of determining whether or not sludge is present in a divided image. Fig. 6 shows a schematic diagram of sludge in divided images D1 to D5 identified in a frame image at a certain point in time. For convenience, the area where sludge is present is indicated by diagonal lines. The learning model 59 can determine whether sludge is present, present, present, present, or absent for the divided images D1 to D5.

[0035] FIG. 7 is a diagram showing an example of a method for calculating the amount of sludge. The amount of sludge can be calculated, for example, by converting it into a numerical value in the range of 0 to 100%. The control unit 51 can calculate the amount of sludge based on the determination result of the presence or absence of sludge in each of the identified divided images. A specific description will be given below. The control unit 51 can calculate the amount of sludge every predetermined time. The predetermined time can be, for example, one minute, but is not limited to this.

[0036] As shown in FIG. 7, the determination cycle for sludge in a divided image is 1 second, and frames at the time of determination are indicated as frame numbers 1 to 60. That is, the presence or absence of sludge in a divided image is determined 60 times per minute. For divided image D1, it is determined that sludge is present 40 times out of 60 determinations made in one minute. Similarly, for divided image D2, it is determined that sludge is present 35 times out of 60 determinations made in one minute, for divided image D3, it is determined that sludge is present 30 times out of 60 determinations made in one minute, for divided image D4, it is determined that sludge is present 15 times out of 60 determinations made in one minute, and for divided image D5, it is determined that sludge is present 10 times out of 60 determinations made in one minute.

[0037] If the number of times that it is determined that sludge is present among 60 judgments made on a divided image in one minute is a predetermined number or more (e.g., 20 times), the judgment result of the presence or absence of sludge in the predetermined time (1 minute) is "present," and if the number of times that it is determined that sludge is present among 60 judgments made on a divided image in one minute is less than the predetermined number, the judgment result of the presence or absence of sludge in the predetermined time (1 minute) is "absent." The judgment results at the predetermined time are "present," "present," "present," "absent," and "absent" for divided images D1 to D5, respectively. That is, the divided images determined to have sludge in the predetermined time are D1, D2, and D3.

[0038] For each divided image (D1, D2, D3) determined to have sludge in a predetermined time, the ratio of the number of times it was determined that sludge was present to the number of times it was determined to have sludge in the predetermined time is calculated. In the example of FIG. 7, divided image D1 is (40 / 60), divided image D2 is (35 / 60), and divided image D3 is (30 / 60). The amount of sludge in a predetermined time can be the ratio of the number of times it was determined that sludge was present to the number of times it was determined to have sludge in all divided images determined to have sludge. In the example of FIG. 7, the amount of sludge is calculated as [(40+35+30) / (60+60+60)], which is 58%. The control unit 51 can calculate the amount of sludge every minute by repeating the same process at each predetermined time.

[0039] As described above, the control unit 51 determines the presence or absence of sludge in each of the divided images (D1 to D5) over a predetermined period of time, and for those divided images (D1 to D3) that are determined to have sludge during that predetermined period of time, the control unit 51 can calculate the amount of sludge during that predetermined period of time based on the ratio of the number of times that it was determined that sludge was present to the number of times that it was determined to have sludge during that predetermined period of time.

[0040] FIG. 8 is a diagram showing an example of the transition of the sludge amount. In FIG. 8, the vertical axis indicates the sludge amount (0 to 100%), and the horizontal axis indicates time. Each plot of the sludge amount is calculated and displayed, for example, every minute. The control unit 51 can output the calculated transition of the sludge amount. The control unit 51 can display the transition of the sludge amount on the display unit 56 or an external display device.

[0041] Conventionally, operators have had to visually check the sludge dryer through a sight glass installed in the sludge dryer at regular intervals (for example, every 30 minutes or every hour). However, according to this embodiment, in contrast to visual checks, the condition of the sludge can be determined continuously (for example, every minute), and the workload of the operator can be reduced.

[0042] Fig. 9 is a diagram showing an example of the relationship between the sludge volume and the moisture content. The relationship between the sludge volume and the moisture content as shown in Fig. 9 can be determined in advance by operating the sludge dryer 30 and collecting data on the sludge volume and the moisture content. In the example of Fig. 9, for convenience, a sludge volume of 20% corresponds to a moisture content of 40%, a sludge volume of 100% corresponds to a moisture content of 60%, and the relationship between the sludge volume and the moisture content is expressed linearly, but the relationship between the sludge volume and the moisture content (numerical examples and linearity) is not limited to the example of Fig. 9.

[0043] The control unit 51 can calculate the moisture content of the sludge based on the calculated sludge volume and the relationship between the sludge volume and the moisture content as shown in Fig. 9. The control unit 51 can output the transition of the calculated moisture content. The control unit 51 can display the transition of the moisture content on the display unit 56 or an external display device. The transition of the moisture content can be represented in the same manner as in Fig. 8.

[0044] According to this embodiment, the moisture content (state) of the sludge during operation of the sludge dryer can be continuously monitored without using a moisture content meter. This makes it possible to stably maintain the moisture content of the sludge (dried sludge) dried by the sludge dryer 53 within a required range (for example, between 40% and 50%).

[0045] In the above-described example, as shown in Fig. 5, the presence or absence of sludge in a segmented image is determined using learning model 59. The presence or absence of sludge in a segmented image is not limited to the configuration using learning model 59. In the following, a method of determining the presence or absence of sludge in a segmented image using image processing unit 54 will be described.

[0046] FIG. 10 is a diagram showing an example of a method for tracking feature points by the image processing unit 54. If the frame rate of the camera 23 is 60 fps, a time series of images of 60 frames per second can be acquired. As shown in FIG. 10, attention is paid to three frames of images, frame (n-1), frame (n), and frame (n+1). The image processing unit 54 extracts feature points representing sludge from the images of each frame. Feature points can be extracted by detecting, for example, edges, corners, parts where color changes, parts where brightness changes, etc. on the image and extracting them as feature points.

[0047] The image processing unit 54 performs matching of feature points by, for example, calculating the distance between feature points extracted from each of two images of adjacent frames and associating feature points with the closest distance. The image processing unit 54 tracks the matched feature points for each frame to track the feature points. The example of Fig. 10 shows how the feature point F1 is tracked across three images, frame (n-1), frame (n), and frame (n+1).

[0048] FIG. 11 is a diagram showing an example of determining the presence or absence of sludge in a divided image based on feature points. FIG. 11 shows a schematic diagram of sludge in divided images D1 to D5 specified in a frame image at a certain time point. Feature points tracked by the image processing unit 54 are represented by dots. The presence or absence of sludge is determined by comparing the feature points extracted in each divided image with a predetermined threshold value, and if the number of feature points is equal to or greater than the threshold value, it can be determined that sludge is "present." If the number of feature points is less than the threshold value, it can be determined that sludge is "absent." In the example of FIG. 11, for divided images D1 to D4, the number of feature points is equal to or greater than the threshold value, so sludge is determined to be "present," and for divided image D5, the number of feature points is less than the threshold value, so sludge is determined to be "absent."

[0049] As described above, the control unit 51 extracts feature points representing sludge based on the identified divided image, and can determine the presence or absence of sludge in the identified divided image based on the number of extracted feature points.The method of calculating the sludge amount is the same as the example in Figure 7, so description will be omitted.

[0050] Next, a case will be described where the state of the sludge is grasped by photographing the area inside the sludge dryer 30 near the midpoint between the inlet 31 and the outlet 32.

[0051] FIG. 12 is a diagram showing an example of the inside of the sludge dryer 30 photographed by the camera 22. FIG. 12 shows an image photographed in the vicinity of the midpoint between the inlet 31 and the outlet 32 ​​inside the sludge dryer 30. In the example of FIG. 12, six heat transfer vanes 34 fixed to one of the rotating shafts 33 are illustrated. In reality, the heat transfer vanes 34 fixed to the other rotating shaft 33 are also photographed at the same time, but for convenience, FIG. 12 illustrates only one of the rotating shafts 33 and the heat transfer vanes 34 fixed to the rotating shaft 33. In the photographed image, divided images D11, D12, D13, D14, and D15 are identified that are arranged in the area between adjacent heat transfer vanes 34 and are aligned along a direction perpendicular to the direction from the rotating shaft 33 to the lower rotating shaft 33.

[0052] That is, the control unit 51 acquires time-series images taken inside the sludge dryer 30 where the sludge is stirred and transported, and identifies divided images D11-D15 obtained by dividing each of the acquired time-series images into a plurality of regions. More specifically, the control unit 51 identifies, as a plurality of divided images, images of regions aligned along a direction perpendicular to the direction of the rotation shaft 33 of the heat transfer blade 34 that stirs and transports the sludge.

[0053] FIG. 13 is a diagram showing an example of judging the presence or absence of sludge in a divided image. FIG. 13 shows sludge in divided images D11 to D15 specified in a frame image at a certain time. Divided image D11 is located near the rotation shaft 33 inside the sludge dryer 30 and is located on the upper side. Divided image D15 is located near the bottom inside the sludge dryer 30 and is located on the lower side. For convenience, the area where sludge exists is shown without a pattern, and the area where sludge exists is shown with diagonal lines. If sludge exists, the color of the sludge is reflected and it looks bright. On the other hand, if sludge does not exist, the dark state inside the sludge dryer 30 is reflected, so the area where sludge does not exist looks dark. In the example of FIG. 13, it is possible to judge whether sludge exists or not for divided images D11 to D15 as "absent", "absent", "present", "present", or "present". That is, in the example of FIG. 13, it is possible to judge to what height sludge exists inside the sludge dryer 30.

[0054] The image processing unit 54 can determine the presence or absence of sludge based on the pixel values ​​of the identified divided image. More specifically, in the identified divided image, for example, a region whose brightness value is equal to or greater than a predetermined brightness threshold can be determined as a region in which sludge exists, and a region whose brightness value is less than the predetermined brightness threshold can be determined as a region in which sludge does not exist.

[0055] This makes it possible to determine the height at which the sludge exists inside the sludge dryer 30. When the sludge exists at a high position inside the sludge dryer 30, it can be determined that the moisture content of the sludge is high, and conversely, when the sludge exists at a low position inside the sludge dryer 30, it can be determined that the moisture content of the sludge is low. This is because the more moisture the sludge contains, the more easily it is lifted and agitated by the heat transfer blades 34.

[0056] In the example of Figure 13, the presence or absence of sludge is determined based on the pixel values ​​of the divided image, but this is not limited to this, and the learning model 59 illustrated in Figure 5 may be used, or a method using feature points illustrated in Figure 10 may be used.

[0057] FIG. 14 is a diagram showing an example of a method for calculating the amount of sludge. The amount of sludge can be calculated, for example, by converting it into a numerical value in the range of 0 to 100%. The control unit 51 can calculate the amount of sludge based on the determination result of the presence or absence of sludge in each of the identified divided images. A specific description will be given below. The control unit 51 can calculate the amount of sludge every predetermined time. The predetermined time can be, for example, one minute, but is not limited to this.

[0058] As shown in FIG. 14, the determination cycle of sludge in a divided image is 1 second, and the frames at the time of determination are indicated by frame numbers 1 to 60. That is, the presence or absence of sludge in a divided image is determined 60 times per minute. The divided image D11 is determined to have sludge 10 times out of 60 determinations made in one minute. Similarly, the divided image D12 is determined to have sludge 40 times out of 60 determinations made in one minute, the divided image D13 is determined to have sludge 45 times out of 60 determinations made in one minute, the divided image D14 is determined to have sludge 60 times out of 60 determinations made in one minute, and the divided image D15 is determined to have sludge 60 times out of 60 determinations made in one minute. Note that the numbers shown in FIG. 14 are merely examples, and are not limited to these.

[0059] If the number of times that it is determined that sludge is present among 60 judgments made on a divided image in one minute is a predetermined number or more (e.g., 20 times), the judgment result of the presence or absence of sludge in the predetermined time (one minute) is "present," and if the number of times that it is determined that sludge is present among 60 judgments made on a divided image in one minute is less than the predetermined number, the judgment result of the presence or absence of sludge in the predetermined time (one minute) is "absent," "present," "present," "present," and "present" for divided images D11 to D15, respectively. That is, the divided images determined to have sludge in the predetermined time are D12, D13, D14, and D15.

[0060] The divided images determined to have sludge at the specified time are the bottom four divided images out of the five divided images (the height of the sludge is the fourth row from the bottom out of the five rows), so the amount of sludge at the specified time is 4 / 5=80%.The control unit 51 repeats the same process at each specified time, thereby calculating the amount of sludge every minute.

[0061] As described above, the control unit 51 determines the presence or absence of sludge in each of the identified divided images (D11 to D15) over a predetermined period of time, and can calculate the amount of sludge in that predetermined period of time based on the ratio of the number of divided images (D12 to D15) that are determined to have sludge in that predetermined period of time to the number of identified divided images.

[0062] Next, a method of continuously grasping the state of the sludge (for example, the amount of sludge, the water content of the sludge, etc.) by photographing the conveyor 13 that transports the sludge discharged from the sludge dryer 30 with the camera 24 will be described.

[0063] Figure 15 shows an image captured by camera 24 of conveyor 13, which transports sludge discharged from sludge dryer 30. As the conveyor chain advances, the conveyor 13 transports the sludge while being pushed by flight bars (flat plates). In Figure 15, the area indicated by the symbol S is the image area for determining whether or not sludge is present.

[0064] The control unit 51 acquires time-series images of the conveyor 13 that transports the sludge discharged from the sludge dryer 30, and determines the presence or absence of sludge from the acquired time-series images. The presence or absence of sludge may be determined, for example, by using a learning model 59 illustrated in Fig. 5, or by a method using feature points illustrated in Fig. 10.

[0065] When the control unit 51 determines that sludge is present in the image, it identifies the color components of the sludge and calculates the water content of the sludge based on the relationship information indicating the correspondence between the color components and the water content that is determined in advance. Note that the image processing unit 54 can be used to identify the color components of the sludge.

[0066] FIG. 16 is a diagram showing an example of relationship information showing the correspondence between color components and moisture content. In FIG. 16, RGB values ​​are used as color components. On the graph shown in FIG. 16, moisture content is associated with RGB values. For example, (R1, G1, B1) corresponds to a moisture content of 40%, (R2, G2, B2) corresponds to a moisture content of 50%, and (R3, G3, B3) corresponds to a moisture content of 60%. Note that the correspondence is merely an example and is not limited to the example in FIG. 16.

[0067] Returning to Fig. 15, in Fig. 15A, R = 113, G = 111, B = 116, and the moisture content of the sludge is relatively high. Also, in Fig. 15B, R = 188, G = 175, B = 141, and the moisture content of the sludge is relatively low. Note that the RGB values ​​are merely an example and are not limited to these values.

[0068] Next, a method of continuously grasping the state of the sludge (for example, the amount of sludge, the water content of the sludge, etc.) by photographing, with a camera 21, the conveyor 12 that transports the sludge fed into the sludge dryer 30 will be described.

[0069] FIG. 17 is a diagram showing an example of the state of sludge dropping from the conveyor 12 that transports the sludge to be fed into the sludge dryer 30 to the distributor 15. FIG. 17A shows a case where the moisture content of the sludge (dewatered sludge) is low, and FIG. 17B shows a case where the moisture content of the sludge (dewatered sludge) is high. As shown in FIG. 17A, when the moisture content of the sludge is low, the adhesion (viscosity) of the sludge is relatively low, so the adhesion between the sludge and the placement surface of the conveyor 12 is low, and the sludge falls at the tip of the conveyor. On the other hand, as shown in FIG. 17B, when the moisture content of the sludge is high, the adhesion (viscosity) of the sludge is relatively high, so the adhesion between the sludge and the placement surface of the conveyor 12 is high, and the sludge falls after passing the tip of the conveyor and wrapping around to the lower side of the conveyor. The applicant has noticed that there is a correlation between the moisture content of the sludge and the drop position.

[0070] Fig. 18 is a diagram showing an example of a method for identifying the position where sludge will fall from the conveyor 12. Fig. 18A shows an image of the tip of the conveyor 12 captured by the camera 21. See Fig. 1 for the positional relationship between the camera 21 and the tip of the conveyor 12. In Fig. 18A, the rectangular area indicated by the symbol P is an image area for identifying the position where the sludge will fall.

[0071] FIG. 18B shows a method for identifying the falling position of sludge. If the frame rate of the camera 21 is 60 fps, then a time series of images of 60 frames per second can be acquired. As shown in FIG. 18B, attention is paid to four frame images P(i), P(i+1), P(i+2), and P(i+3) of frame (i), frame (i+1), frame (i+2), and frame (i+3). The image processing unit 54 extracts feature points representing sludge from the image of each frame. The feature points can be extracted, for example, by detecting edges, corners, parts where color changes, parts where brightness changes, and the like on the image.

[0072] The image processing unit 54 performs matching of feature points by, for example, calculating the distance between feature points extracted from each of two images of adjacent frames and associating feature points with the closest distance. The image processing unit 54 tracks the matched feature points for each frame to track the feature points. The example of FIG. 18B shows how the feature point F1 is tracked across four images, frame (i), frame (i+1), frame (i+2), and frame (i+3).

[0073] When the moving distance of the feature point F1 between the images of adjacent frames is longer than a predetermined threshold, the control unit 51 can determine that sludge has fallen in the frame, and can identify the falling position of the sludge based on the position of the feature point in the image of the frame immediately adjacent to the frame. Note that in the example of FIG. 18B, for convenience, only one feature point is illustrated, but in reality, multiple feature points can be tracked. In this case, the presence or absence of falling can be determined based on the statistical value (e.g., average value, median value, etc.) of the moving distance of each feature point, and the falling position can be represented, for example, by the center of gravity position of the multiple feature points. In the example of FIG. 18B, it can be determined that sludge has fallen in frame (i+3), and the falling position of the sludge can be identified in the image of frame (i+2).

[0074] Fig. 19 is a diagram showing the relationship between the falling position of sludge and the adhesion degree or moisture content. As shown in Fig. 19, the falling position of sludge at the front end of conveyor 12 is represented by Y1, and the position where sludge falls after passing the front end of conveyor 12 and wrapping around to the lower side of conveyor 12 is represented by Y2. Since the adhesion degree of sludge falling at fall position Y1 is low and the adhesion degree of sludge falling at fall position Y2 is high, the correspondence relationship between the falling position and the adhesion degree can be expressed as shown in Fig. 19. Similarly, the moisture content of sludge falling at fall position Y1 is low and the moisture content of sludge falling at fall position Y2 is high, and the correspondence relationship between the falling position and the moisture content can be expressed as shown in Fig. 19.

[0075] As described above, the control unit 51 acquires time-series images of the conveyor 12 transporting the sludge to be fed into the sludge dryer 30, identifies the fall position of the sludge on the conveyor 12 based on the acquired time-series images, and calculates the moisture content of the sludge based on the identified fall position.

[0076] Furthermore, the control unit 51 can extract feature points representing sludge based on the acquired time-series images, and identify the falling position where the sludge will fall based on the relative position between a first feature point extracted from the image at a first time point and a second feature point corresponding to the first feature point extracted from the image at a second time point after the first time point.

[0077] Furthermore, the control unit 51 can calculate the degree of adhesion of sludge based on the identified drop position.

[0078] Furthermore, by using the method shown in FIG. 16, the control unit 51 can calculate the water content of the sludge based on pixel values ​​(for example, color components such as RGB) of the acquired time-series images.

[0079] Next, an example of the display of the quantification of the state of sludge will be described.

[0080] FIG. 20 is a diagram showing an example of a display of the transition of the sludge amount inside the sludge dryer 30. In FIG. 20, the upper part displays the transition of the sludge amount at the sludge dryer midpoint based on images taken by the camera 22, and the lower part displays the transition of the sludge amount at the sludge dryer outlet side based on images taken by the camera 23. Upper and lower limit values ​​are set for the transition of the sludge amount at the sludge dryer midpoint and the sludge dryer outlet side, respectively, and the set upper and lower limit values ​​are displayed together with the transition of the sludge amount. When the transition of the sludge amount exceeds the upper limit value or falls below the lower limit value, an alarm is output. In the example of FIG. 20, an alarm is output at the timing when the sludge amount at the sludge dryer midpoint exceeds the upper limit value.

[0081] As described above, the control unit 51 can output an alarm when the calculated sludge amount falls outside the normal range. This allows a quick response to stabilize the operation of the sludge dryer 30.

[0082] FIG. 21 is a diagram showing an example of display of the transition of the moisture content of the sludge. In FIG. 21, the upper left side displays the transition of the moisture content of the sludge at the input side of the sludge dryer based on the image taken by the camera 21, the lower side displays the transition of the moisture content of the sludge inside the sludge dryer based on the images taken by the cameras 22 and 23, and the upper right side displays the transition of the moisture content of the sludge at the discharge side of the sludge dryer based on the image taken by the camera 24. An upper limit and a lower limit are set for the moisture content of the sludge at the discharge side of the sludge dryer, and the set upper limit and lower limit are displayed together with the transition of the moisture content. When the transition of the moisture content exceeds the upper limit or falls below the lower limit, an alarm is output. Although not shown, upper limit and lower limit may be displayed for the transition of the moisture content of the sludge at the input side of the sludge dryer and the transition of the moisture content of the sludge inside the sludge dryer.

[0083] As described above, the control unit 51 can output an alarm when the calculated moisture content is outside the normal range. This allows a quick response to stabilize the operation of the sludge dryer 30.

[0084] FIG. 22 is a diagram showing an example of the operation control screen of the sludge dryer 30. In the operation control screen, for example, the upper part displays the transition of the moisture content of the sludge on the discharge side of the sludge dryer based on the image taken by the camera 24, and the lower part displays the transition of the amount of sludge inside the sludge dryer based on the image taken by the camera 22 or the camera 23. As shown in FIG. 22, at time t1, the moisture content of the sludge on the discharge side of the sludge dryer 30 reaches the upper limit value, the sludge condition is in an over-humid state, and the amount of sludge inside the sludge dryer 30 is small but within the normal range. In such a case, the control unit 51 can output a control signal to the sludge dryer controller 40 to reduce the amount of sludge processed. This can eliminate the over-humidity. However, on the other hand, the amount of sludge input to the sludge dryer 30 may further decrease and fall outside the normal range. Therefore, in such a case, the control unit 51 outputs a control signal to the sludge dryer controller 40 to increase the steam pressure. This makes it possible to prevent a future alarm from being issued due to the sludge amount being outside the normal range, while eliminating the overly wet condition without changing the sludge amount.

[0085] As described above, the control unit 51 can control operating parameters including at least one of the sludge processing volume and steam pressure of the sludge dryer 30 based on the amount of sludge inside the sludge dryer 30 and the moisture content of the sludge discharged from the sludge dryer 30.

[0086] 23 is a diagram showing a first example of a processing procedure by the operation support device 50. In the following explanation, for convenience, the processing will be performed by the control unit 51. The control unit 51 acquires time-series images of the conveyor 12 that transports the sludge to be input into the sludge dryer 30 (S11), and identifies the position where the sludge on the conveyor 12 will fall based on the acquired time-series images (S12).

[0087] The control unit 51 calculates the moisture content of the sludge based on the dropped position (S13), outputs the calculated change in moisture content (S14), and ends the process.

[0088] 24 is a diagram showing a second example of a processing procedure by the operation support device 50. The control unit 51 acquires time-series images of the inside of the sludge dryer 30 (S21), and identifies divided images obtained by dividing each of the acquired time-series images into a plurality of regions (S22). The control unit 51 determines the presence or absence of sludge in each of the identified divided images (S23).

[0089] The control unit 51 calculates the proportion of sludge present within a predetermined time period for each divided image (S24). The control unit 51 determines the presence or absence of sludge for each divided image for a predetermined time unit based on the calculated proportion (S25), and calculates the amount of sludge based on the proportion of sludge present within the predetermined time period for divided images determined to have sludge present in the predetermined time unit (the proportion of the number of times it was determined that sludge was present to the number of times it was determined to have sludge present) (S26).

[0090] The control unit 51 calculates the moisture content of the sludge based on the calculated sludge volume (S27), outputs the calculated sludge volume and the transition of the moisture content (S28), and ends the process.

[0091] 25 is a diagram showing a third example of a processing procedure by the operation support device 50. The control unit 51 acquires time-series images of the conveyor 13 that transports the sludge discharged from the sludge dryer 30 (S31), and determines the presence or absence of sludge from the acquired time-series images (S32). If it is determined that there is no sludge (NO in S33), the control unit 51 continues the processing from step S31 onwards.

[0092] If it is determined that sludge is present (YES in S33), the control unit 51 identifies the color components of the sludge (S34), and calculates the moisture content of the sludge using a predetermined relationship between the color components and the moisture content (S35). The control unit 51 outputs the calculated change in moisture content (S36), and ends the process.

[0093] As described above, according to this embodiment, the operation of the sludge dryer 30 can be supported using time-series images of the sludge inside the sludge dryer 30, the sludge (dewatered sludge) fed into the sludge dryer 30, and the sludge (dried sludge) discharged from the sludge dryer 30, so that the moisture content of the sludge (dried sludge) can be stabilized within a required range and energy-saving operation of the sludge dryer 30 is possible. In addition, the sludge condition, such as the sludge volume and moisture content of the sludge, can be quantified and the sludge condition can be continuously monitored, so that the operator's judgment of condition monitoring can be standardized and manpower can be reduced since visual monitoring by the operator is not required.

[0094] Fig. 26 is a diagram showing a second example of the configuration of the driving support device 50. The difference from the first example of the driving support device shown in Fig. 3 is that a learning model 60 is provided. Although not shown, the second example may also include a learning model 59. Details of the learning model 60 will be described later. Also, the same parts as in the first example are given the same reference numerals, and description thereof will be omitted.

[0095] Figure 27 shows an image captured by camera 24 of conveyor 13, which transports sludge discharged from sludge dryer 30. As the conveyor chain advances, flight bars (flat plates) of conveyor 13 push the sludge while transporting it. The image shown in Figure 27 is the same as the image shown in Figure 15.

[0096] The control unit 51 acquires time-series images of the conveyor transporting the sludge. More specifically, the control unit 51 acquires time-series images of the conveyor 13 transporting the sludge discharged from the sludge dryer 30, taken by the camera 24. The control unit 51 can identify the state of the sludge based on the acquired time-series images. A method for identifying the state of the sludge will be described below.

[0097] The control unit 51 performs pre-processing of the acquired time-series images for image analysis. In the pre-processing, the average brightness of the upper surface of the flight bar (also called "flight") is calculated by utilizing the fact that the upper surface of the flight bar reflects the light of the illumination and appears bright, and the image at the time when the brightness is the highest during the period (for example, about 7 to 8 seconds) in which the flight bar moves is extracted and the extracted image is acquired. That is, the time-series images can be acquired at each movement period of the flight bar. Also, an image including an area necessary for image analysis from the entire area of ​​the image of the conveyor 13 can be extracted and the extracted image can be acquired. The area necessary for image analysis may be fixedly determined, but may be changed as appropriate depending on the conveyor or transport system that transports the sludge. This improves the accuracy of identifying the state of the sludge.

[0098] FIG. 28 is a diagram showing an example of processing by the learning model 60. As shown in FIG. 28, when an image of sludge is input to the learning model 60, the learning model 60 outputs a class. In the example of FIG. 28, one or more classes out of 30 classes, classes 1 to 30, can be output. The classes are classes for classifying the state of sludge, and for example, classes can be determined according to the particle size of the sludge. Note that the number of classes is not limited to 30.

[0099] The learning model 60 can be configured with a convolutional neural network (CNN) specialized for image analysis, but models such as YOLO (You Only Look Once) and SSD (single shot multibox detector) may also be used.

[0100] The learning model 60 can be trained (generated) using, for example, "unsupervised learning". In unsupervised learning, a large amount of images are input to the learning model 60, which extracts features of the state of the sludge (e.g., particle size of the sludge) from the input images, and performs class classification by clustering images with close or similar extracted features. In this way, when an image is input, the learning model 60 can extract features of the input image and output a class corresponding to the input image. Furthermore, the learning model 60 can output the accuracy (probability) of the class to be output. Furthermore, when there are multiple classes that can be classified, the learning model 60 can output multiple classes. In this case, the class with the highest accuracy can be set as the output value.

[0101] FIG. 29 is a diagram showing an example of classified images. The learning model 60 extracts features of the sludge state from the input image to perform class classification, and as shown in FIG. 29, when images of different classes are compared, images with close or similar sludge states (e.g., sludge particle size) are included. For example, classes 2, 6, and 11 shown in FIG. 29 include images with close or similar particle sizes. This is thought to be because, when the learning model 60 extracts features, it extracts not only the particle size of the sludge but also the amount of sludge and the area (size) of the sludge as features. Therefore, the classes are converted into a scale according to the particle size of the sludge.

[0102] FIG. 30 is a diagram showing an example of the correspondence relationship between classes and scales. The scales are classified according to the particle size of the sludge (dried sludge). The scales are classified into, for example, 0 to 4, and scale 0 indicates a case where there is no sludge (i.e., no dried sludge is transported to the flight bar portion). In addition, the larger the value of scales 1 to 4, the larger the particle size. The correspondence relationship between classes and scales can be set in advance and stored in the storage unit 57. Scale 0 corresponds to one or more classes among classes 1 to 30 that do not contain sludge. Scale 1 corresponds to one or more classes among classes 1 to 30 that have small sludge particle sizes, excluding the class corresponding to scale 0. Scale 2 corresponds to one or more classes among classes 1 to 30 that have larger sludge particle sizes than scale 1, excluding the classes corresponding to scale 0 and scale 1. Scale 3 corresponds to one or more classes among classes 1 to 30 that have larger sludge particle sizes than scale 2, excluding the classes corresponding to scales 0 to 2. Scale 4 corresponds to one or more classes among classes 1 to 30, excluding classes corresponding to scales 0 to 3, in which the sludge particle size is larger than that of scale 3. The correspondence between classes and scales may be determined manually according to the particle size of the sludge. The number of scales is not limited to the example in FIG.

[0103] FIG. 31 is a diagram showing an example of an image of sludge at each scale. As shown in FIG. 31, at scale 1, the particle size of the sludge (dried sludge) is small. At scale 2, the particle size of the sludge is larger than at scale 1. At scale 3, the particle size of the sludge is even larger, and at scale 4, the particle size of the sludge is even larger than at scale 3. By using the scales, it is possible to identify the particle size as the state of the sludge.

[0104] The control unit 51 can identify the particle size of the sludge based on the acquired time-series images, and identify the identified particle size as the state of the sludge. Specifically, the control unit 51 inputs the acquired time-series images to the learning model 60. For each input image, the learning model 60 outputs a class according to the features of the sludge contained in the image. The control unit 51 converts the class output by the learning model 60 into a scale.

[0105] FIG. 32 is a diagram showing an example of the state of sludge. In FIG. 32, the vertical axis indicates the scale (i.e., the particle size of the sludge), and the horizontal axis indicates the time. The example in FIG. 32 shows the time transition of the scale converted by the control unit 51 based on a time series of images taken by the camera 24 of the conveyor 13 that transports the sludge discharged from the sludge dryer 30. Specifically, the figure shows the transition of the moving average of the scale values ​​(1 to 4) for about one minute (about eight images). Note that data of scale 0, which indicates no sludge, is excluded. In this way, the control unit 51 can output the time transition of the scale (particle size) as the state of the sludge.

[0106] FIG. 33 is a diagram showing an example of the correspondence relationship between scale and the moisture content of sludge. FIG. 33 is a schematic representation of the correspondence relationship between scale (particle size of sludge) and the moisture content of sludge, with the larger the particle size of the sludge, the higher the moisture content. Correspondence information showing the correspondence relationship between scale and the moisture content of sludge can be stored in the memory unit 57. The control unit 51 can specify the moisture content of the sludge as the state of the sludge based on the specified scale (particle size of sludge) and the correspondence information. Furthermore, the control unit 51 can output the time transition of the moisture content as the state of the sludge.

[0107] Specifically, as illustrated in FIG. 21 (the chart on the upper right side), the transition of the moisture content of the sludge at the discharge side of the sludge dryer can be displayed based on images captured by the camera 24. That is, the control unit 51 can output the transition of the calculated moisture content. The control unit 51 can display the transition of the moisture content on the display unit 56 or an external display device. Furthermore, the control unit 51 can output an alarm when the calculated moisture content falls outside the normal range. This allows prompt action to be taken to stabilize the operation of the sludge dryer 30.

[0108] 34 is a diagram showing a fourth example of a processing procedure by the operation support device 50. The control unit 51 acquires a time series of images of the conveyor 13 that transports the sludge discharged from the sludge dryer 30 (S41), and identifies a class of the acquired time series of images (S42). The control unit 51 converts the identified class into a scale (S43), and calculates the moisture content of the sludge using a predetermined relationship between the scale and the moisture content (S44). The control unit 51 outputs the change in the moisture content (S45), and ends the processing.

[0109] (Appendix 1) The computer program acquires a time series of images taken inside the sludge dryer where the sludge is stirred and transported, and divides each of the acquired time series images into multiple regions to create segmented images. The present invention causes a computer to execute a process of: determining whether or not sludge is present in each of the identified divided images; and calculating the amount of sludge based on the determination result of the presence or absence of sludge.

[0110] (Appendix 2) The computer program in Appendix 1 causes a computer to execute a process of determining the presence or absence of sludge in each of the divided images over a predetermined time period, and for those divided images determined to have sludge during the predetermined time period, calculating the amount of sludge during the predetermined time period based on the ratio of the number of times that the images were determined to have sludge to the number of times that the images were determined to have sludge during the predetermined time period.

[0111] (Supplementary Note 3) The computer program causes a computer to execute a process of calculating the moisture content of the sludge based on the calculated sludge volume in Supplementary Note 1 or Supplementary Note 2.

[0112] (Appendix 4) In any one of Appendices 1 to 3, the computer program causes a computer to execute a process in which, when a segmented image taken inside a sludge dryer is input, the identified segmented image is input into a learning model that determines the presence or absence of sludge, and the presence or absence of sludge in the identified segmented image is determined.

[0113] (Appendix 5) The computer program in any one of Appendices 1 to 4 causes a computer to execute a process of extracting feature points representing sludge based on a specified divided image, and determining the presence or absence of sludge in the specified divided image based on the number of extracted feature points.

[0114] (Supplementary Note 6) The computer program causes a computer to execute a process of identifying, as a divided image, an image of a region between heat transfer vanes that mix and transport sludge in Supplementary Note 4 or Supplementary Note 5.

[0115] (Supplementary Note 7) The computer program according to any one of Supplementary Note 1 to Supplementary Note 6 causes a computer to execute a process of determining the presence or absence of sludge based on pixel values ​​of the identified divided image.

[0116] (Appendix 8) The computer program in Appendix 7 causes a computer to execute a process of identifying, as a plurality of divided images, images of areas aligned along a direction perpendicular to the direction of the rotation axis of the heat transfer blade that mixes and transports the sludge.

[0117] (Appendix 9) The computer program in any one of Appendices 1 to 8 causes a computer to execute the following process: acquire time-series images of a conveyor that transports sludge to be fed into a sludge dryer, identify a position where the sludge on the conveyor will fall based on the acquired time-series images, and calculate the moisture content of the sludge based on the identified position.

[0118] (Supplementary Note 10) The computer program in Supplementary Note 9 causes a computer to execute a process of extracting feature points representing sludge based on the acquired time-series images, and identifying a falling position at which the sludge will fall based on a relative position between a first feature point extracted from an image at a first time point and a second feature point corresponding to the first feature point extracted from an image at a second time point that is later than the first time point.

[0119] (Supplementary Note 11) The computer program causes a computer to execute a process of calculating the degree of adhesion of sludge based on the identified falling position in Supplementary Note 9 or Supplementary Note 10.

[0120] (Supplementary Note 12) The computer program according to any one of Supplements 9 to 11 causes a computer to execute a process of calculating the moisture content of sludge based on pixel values ​​of the acquired time-series images.

[0121] (Appendix 13) The computer program according to any one of Appendices 1 to 12 acquires a time series of images of a conveyor that transports sludge discharged from a sludge dryer, and The computer executes a process to determine the presence or absence of sludge from the acquired time-series images, and if it is determined that sludge is present in the images, to identify the color components of the sludge and calculate the moisture content of the sludge based on relationship information that indicates the correspondence between predetermined color components and the moisture content.

[0122] (Appendix 14) The computer program causes a computer to execute a process of outputting the calculated change in sludge volume in any one of appendix 1, appendix 2, appendix 4, appendix 5, appendix 7, or appendix 8.

[0123] (Appendix 15) The computer program causes a computer to execute a process of outputting an alarm when the calculated sludge amount falls outside the normal range in any one of Appendix 1, Appendix 2, Appendix 4, Appendix 5, Appendix 7, or Appendix 8.

[0124] (Appendix 16) The computer program in appendix 13 causes a computer to execute a process of controlling operating parameters including at least one of a sludge processing volume and a steam pressure of the sludge dryer based on the volume of sludge inside the sludge dryer and the moisture content of the sludge discharged from the sludge dryer.

[0125] (Appendix 17) The computer program causes a computer to execute a process of acquiring a time series of images of a conveyor transporting sludge, and identifying the state of the sludge based on the acquired time series of images.

[0126] (Supplementary Note 18) The computer program in Supplementary Note 17 causes a computer to execute a process of identifying a particle size of sludge based on the acquired time-series images, and identifying the identified particle size as a state of the sludge.

[0127] (Appendix 19) The computer program in appendix 18 causes a computer to execute a process of storing correspondence information indicating a correspondence between sludge particle size and moisture content in advance, and identifying the moisture content corresponding to the identified particle size as the state of the sludge.

[0128] (Supplementary Note 20) The computer program causes a computer to execute a process of outputting the transition of the moisture content in any one of Supplementary Note 3, Supplementary Note 9 to Supplementary Note 11, Supplementary Note 13, or Supplementary Note 19.

[0129] (Additional Note 21) The computer program causes a computer to execute a process of outputting an alarm when the moisture content falls outside a normal range in any one of Additional Note 3, Additional Note 9 to Additional Note 11, Additional Note 13, or Additional Note 19.

[0130] (Appendix 22) The operation assistance device is an operation assistance device for a sludge dryer, and is equipped with a control unit, which acquires a time series of images taken inside the sludge dryer where sludge is stirred and transported, identifies divided images divided into a plurality of areas in each of the acquired time series images, determines the presence or absence of sludge in each of the identified divided images, and calculates the amount of sludge based on the determination result of the presence or absence of sludge.

[0131] (Appendix 23) The operation assistance device is an operation assistance device for a sludge dryer, and is equipped with a control unit, which acquires time-series images of a conveyor that transports sludge, and identifies the condition of the sludge based on the acquired time-series images.

[0132] (Appendix 24) The operation assistance system includes the above-mentioned operation assistance device, a sludge dryer, a first photographing device that photographs the inside of the sludge dryer, a second photographing device that photographs the conveyor that transports sludge fed into the sludge dryer, and a third photographing device that photographs the conveyor that transports sludge discharged from the sludge dryer.

[0133] (Appendix 25) The operation support method is an operation support method for a sludge dryer, which acquires a time series of images taken inside the sludge dryer where sludge is stirred and transported, identifies divided images divided into a plurality of areas in each of the acquired time series images, determines whether or not sludge is present in each of the identified divided images, and calculates the amount of sludge based on the results of the determination of whether or not sludge is present.

[0134] (Appendix 26) The operation support method is a method for supporting operation of a sludge dryer, which involves obtaining time-series images of a conveyor that transports sludge, and identifying the condition of the sludge based on the obtained time-series images.

[0135] The matters described in each embodiment can be combined with each other. In addition, the independent claims and dependent claims described in the claims can be combined with each other in any and all combinations regardless of the citation format. Furthermore, the claims use a format in which a claim cites two or more other claims (multiple claim format), but this is not limited to this. A multiple claim (multi-multi claim) that cites at least one multiple claim may also be used. [Explanation of symbols]

[0136] 10 Dehydrator 11, 12, 13 Conveyor 15 Distributor 20 Incinerator 21, 22, 23, 24 Camera 30 Sludge dryer 31 Inlet 32, 25 outlet 33 Rotation axis 34 Heat transfer blade 35 Heat medium pipe 50 Driving assistance device 51 Control section 52 Communications Department 53 Memory 54 Image processing section 55 Interface section 56 Display section 57 Memory section 58 Computer Programs 59, 60 Learning Model

Claims

1. A time series of images was acquired of the inside of the sludge dryer where the sludge was being stirred and transported. Identifying divided images into multiple regions for each of the acquired time series images; The presence or absence of sludge in each of the identified divided images is determined; Calculating the amount of sludge based on the result of determining whether or not sludge is present; A computer program that causes a computer to carry out processing.

2. determining the presence or absence of sludge in each of the divided images over a predetermined period of time; calculating an amount of sludge for the predetermined time period based on a ratio of the number of times that it was determined that sludge was present for the divided images determined to have sludge present for the predetermined time period to the number of times that it was determined to have sludge present for the divided images determined to have sludge present for the predetermined time period; 2. A computer program product according to claim 1, which causes a computer to carry out a process.

3. Calculating the moisture content of the sludge based on the calculated sludge volume; 2. A computer program product according to claim 1, which causes a computer to carry out a process.

4. When a segmented image of the inside of a sludge dryer is input, the specified segmented image is input to a learning model for determining the presence or absence of sludge, and the presence or absence of sludge in the specified segmented image is determined.

2. A computer program product according to claim 1, which causes a computer to carry out a process.

5. Extracting feature points representing sludge based on the identified divided images; determining whether or not sludge is present in the specified divided image based on the number of extracted feature points; 2. A computer program product according to claim 1, which causes a computer to carry out a process.

6. Identifying an image of a region between heat transfer blades that mix and transport sludge as a segmented image; 6. A computer program product according to claim 4 or 5, which causes a computer to execute a process.

7. determining whether or not sludge is present based on pixel values ​​of the identified divided image; 2. A computer program product according to claim 1, which causes a computer to carry out a process.

8. Identifying, as a plurality of divided images, images of areas aligned along a direction perpendicular to the direction of a rotation axis of a heat transfer blade that mixes and transports the sludge; 8. A computer program product according to claim 7, which causes a computer to carry out a process.

9. A time series of images was acquired of the conveyor that transports the sludge to be fed into the sludge dryer. Identifying a position where the sludge on the conveyor will fall based on the acquired time-series images; Calculating the moisture content of the sludge based on the identified falling position; 2. A computer program product according to claim 1, which causes a computer to carry out a process.

10. Extracting feature points representing sludge based on the acquired time series of images; identifying a falling position where the sludge will fall based on a relative position between a first feature point extracted from an image at a first time point and a second feature point corresponding to the first feature point extracted from an image at a second time point that is later than the first time point; 10. A computer program product according to claim 9, which causes a computer to carry out a process.

11. Calculating the degree of adhesion of sludge based on the identified falling position; 10. A computer program product according to claim 9, which causes a computer to carry out a process.

12. Calculate the moisture content of the sludge based on the pixel values ​​of the acquired time series images. A computer program product according to any one of claims 9 to 11, which causes a computer to carry out a process.

13. A time series of images was acquired of the conveyor that transports the sludge discharged from the sludge dryer. The presence or absence of sludge is determined from the acquired time series images. If it is determined that sludge is present in the image, a color component of the sludge is identified; Calculating the moisture content of the sludge based on relationship information indicating a correspondence relationship between a predetermined color component and the moisture content; 2. A computer program product according to claim 1, which causes a computer to carry out a process.

14. Output the calculated sludge volume trend, A computer program product according to any one of claims 1 to 8, which causes a computer to carry out a process.

15. If the calculated sludge volume is outside the normal range, an alarm is issued. A computer program product according to any one of claims 1 to 8, which causes a computer to carry out a process.

16. controlling operation parameters including at least one of a sludge throughput and a steam pressure of the sludge dryer based on the amount of sludge in the sludge dryer and the moisture content of the sludge discharged from the sludge dryer; 14. A computer program product according to claim 13, which causes a computer to carry out a process.

17. A time series of images of the conveyor transporting the sludge was acquired. Identifying the condition of the sludge based on the acquired time series of images; A computer program that causes a computer to carry out processing.

18. Based on the acquired time series images, the particle size of the sludge is identified, The specified particle size is specified as the state of the sludge.

20. A computer program product according to claim 17, which causes a computer to carry out a process.

19. Correspondence information showing a correspondence relationship between the particle size and the moisture content of sludge is stored in advance; The moisture content corresponding to the identified particle size is identified as the state of the sludge.

20. A computer program product according to claim 18, which causes a computer to carry out a process.

20. Outputting the transition of the moisture content. A computer program product according to any one of claims 3, 9 to 11, 13 and 19, which causes a computer to carry out a process.

21. If the moisture content falls outside the normal range, an alarm is output. A computer program product according to any one of claims 3, 9 to 11, 13 and 19, which causes a computer to carry out a process.

22. An operation support device for a sludge dryer, comprising: A control unit is provided, The control unit is A time series of images was acquired of the inside of the sludge dryer where the sludge was being stirred and transported. Identifying divided images into multiple regions for each of the acquired time series images; The presence or absence of sludge in each of the identified divided images is determined; Calculating the amount of sludge based on the result of determining whether or not sludge is present; Driving assistance device.

23. An operation support device for a sludge dryer, comprising: A control unit is provided, The control unit is A time series of images of the conveyor transporting the sludge was acquired. Identifying the condition of the sludge based on the acquired time series of images; Driving assistance device.

24. A driving assistance device according to claim 22 or 23; A sludge dryer, A first photographing device for photographing the inside of the sludge dryer; A second photographing device that photographs a conveyor that transports the sludge to be fed into the sludge dryer; a third imaging device for imaging a conveyor that transports the sludge discharged from the sludge dryer; A driving assistance system equipped with

25. A method for supporting operation of a sludge dryer, comprising: A time series of images was acquired of the inside of the sludge dryer where the sludge was being stirred and transported. Identifying divided images into multiple regions for each of the acquired time series images; The presence or absence of sludge in each of the identified divided images is determined; Calculating the amount of sludge based on the result of determining whether or not sludge is present; Driving assistance methods.

26. A method for supporting operation of a sludge dryer, comprising: A time series of images of the conveyor transporting the sludge was acquired. Identifying the condition of the sludge based on the acquired time series of images; Driving assistance methods.

Images