System for estimating the chip generation area

Abstract

This invention provides an estimation system that can acquire clear image data of the vicinity of an object being cut using a cutting machine, and can estimate the chip generation area based solely on the acquired image data. [Solution] The estimation system 100 of the present invention is a system for estimating the chip generation area generated by cutting an object using a cutting machine. The estimation system 100 includes an image acquisition device 10 including a housing 13 having a hollow housing 11 and a rotating window 12 provided in an opening of the housing 11, a lens 14 positioned in the center of the housing 11 so as to face the rotating window 12, and a camera 15, and an estimation device 20 that estimates the chip generation area based on image data IDs of the vicinity of the object acquired by the image acquisition device 10.

Description

[Technical Field]

[0001] The present invention relates to a system for estimating the chip generation area, and more particularly to a system for estimating the chip generation area generated by cutting an object using a cutting machine. [Background technology]

[0002] When machining a workpiece using a cutting machine, a large amount of chips are generated. If machining continues without addressing the generated chips, the chips can damage the workpiece and affect its accuracy. Furthermore, excessive chip generation can be an indicator of damage to the cutting machine's cutting edge or errors in the cutting parameter settings. For this reason, various devices or systems have been proposed to estimate the chip generation area from the workpiece.

[0003] Incidentally, when machining an object using a cutting machine, a large amount of water, coolant, oil, etc., is scattered around, making it difficult to see the area where chips are generated. Therefore, there is a particular need for a device or system that can estimate the area where chips are generated from the object, regardless of whether or not visibility is impaired by a large amount of water, coolant, oil, etc.

[0004] For example, Patent Document 1 describes an estimation device that estimates the chip accumulation region, which is the area where chips accumulate, regardless of whether or not poor visibility occurs. In Patent Document 1, if the captured image acquired by the image acquisition unit shows that chips are visible, the chip accumulation region is estimated based on the captured image. If the captured image acquired by the image acquisition unit shows that chips are not visible, the chip accumulation region is estimated based on cutting information. Captured image refers to an image of chips generated from a workpiece cut by a processing machine. Cutting information refers to information about the workpiece being cut by the processing machine. [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] Patent No. 7483162 [Overview of the Initiative] [Problems that the invention aims to solve]

[0006] In Patent Document 1, in order to estimate the chip accumulation area, it is necessary to determine the visibility of the chips in the captured video, and the information used to estimate the chip accumulation area differs depending on the determined visibility. The technology described in Patent Document 1 has room for reconsideration because it requires a complex process and a complex equipment configuration to realize it.

[0007] The present invention has been made in view of these points, and aims to provide an estimation system that can acquire clear image data of the vicinity of an object being cut using a cutting machine, regardless of the environment in which large amounts of water, coolant, oil, etc. are scattered, and can estimate the chip generation area based solely on the acquired image data. [Means for solving the problem]

[0008] To achieve the above objective, the present invention provides a system for estimating the chip generation area generated by cutting an object using a cutting machine, comprising: a housing having a hollow housing and a rotating window provided in an opening of the housing; an image acquisition device including a lens and a camera positioned in the center of the housing so as to face the rotating window; and an estimation device for estimating the chip generation area based on image data of the vicinity of the object acquired by the image acquisition device, wherein the rotating window has a light-transmitting window, a window frame that supports the window at its peripheral edge, and a driven means provided on the outer circumference of the window frame; and the image acquisition device is provided with a driving means for rotating the rotating window by driving the driven means to rotate the window frame, and an injection means for injecting fluid toward the outer surface of the window, thereby providing a chip generation area estimation system (Invention 1).

[0009] According to this invention (Invention 1), clear image data of the vicinity of an object being cut using a cutting machine can be obtained regardless of the environment in which large amounts of water, coolant, oil, etc. are splashed, and the chip generation area can be estimated based solely on the acquired image data.

[0010] In the above invention (Invention 1), the estimation device may include an extracted image creation unit that creates a plurality of extracted images specifying the chip generation region for each of the plurality of still images constituting the image data acquired by the image acquisition device, and a determination unit that determines whether or not the area of ​​the chip generation region exceeds a threshold for each of the plurality of extracted images (Invention 2).

[0011] According to this invention (Invention 2), for example, when the determination unit determines that the area of ​​the chip generation region for a particular extracted image exceeds a threshold, it is possible to alert the user with alarm information.

[0012] In the above invention (Invention 2), when the determination unit detects that the area of ​​the chip generation region exceeds a threshold value in the extracted image, it may output alarm information to an external display device or the cutting machine (Invention 3).

[0013] According to this invention (Invention 3), for example, by checking a display device that shows alarm information, measures such as stopping the operation of the cutting machine and removing chips can be taken. For example, the operation can be controlled in advance so that the cutting machine automatically stops when it receives alarm information.

[0014] In the above inventions (Inventions 2 and 3), the estimation system further comprises a storage device having a storage unit for temporarily storing the image data, and when the determination unit detects an extracted image in which the area of ​​the chip generation region exceeds a threshold, the storage device may transmit the image data before and after the threshold exceedance, which is temporarily stored in the storage unit, to the extracted image creation unit (Invention 4).

[0015] According to such an invention (Invention 4), for example, the estimation accuracy of the generation region of cutting powder by the estimation device can be improved.

[0016] In the above invention (Invention 4), the image data before and after exceeding the threshold value includes first image data previously acquired by the image acquisition device and second image data acquired in real time by the image acquisition device. The extraction image creation unit creates a plurality of first extraction images in which the generation region of cutting powder is specified for each of a plurality of first still images constituting the first image data. The determination unit classifies the plurality of first still images into first learning images with cutting powder and first non-learning images without cutting powder, and uses a learning model learned with the first learning images, the plurality of first extraction images, and the first non-learning images as first teacher data to estimate the generation region of cutting powder (Invention 5).

[0017] According to such an invention (Invention 5), the generation region of cutting powder can be estimated simply and accurately in real time.

[0018] In the above invention (Invention 5), the extraction image creation unit creates a plurality of second extraction images in which the generation region of cutting powder is specified for each of a plurality of second still images constituting the second image data. The determination unit classifies the plurality of second still images into second learning images with cutting powder and second non-learning images without cutting powder, and uses the learning model further learned with the second learning images as second teacher data to estimate the generation region of cutting powder (Invention 6).

[0019] According to such an invention (Invention 6), for example, the second learning image acquired during the present operation can be fed back, so that the generation region of cutting powder can be estimated more accurately in real time.

[0020] In the above inventions (Inventions 1-6), the cutting machine is installed in a chamber, and the image acquisition device may be provided with an attachment portion capable of attaching the image acquisition device to the inner wall of the chamber (Invention 7).

[0021] According to such an invention (Invention 7), the image acquisition device can be attached to an optimal position where it is easier to acquire image data near the object by the attachment part.

[0022] In the above invention (Inventions 1 - 7), the housing may further include supply means for supplying gas into the housing, and the injection means may be configured to inject the gas supplied by the supply means toward the outer surface of the window (Invention 8).

[0023] According to such an invention (Invention 8), by using the gas supplied from the supply means, there is no need to separately supply gas to the injection means, and the cost can be reduced.

[0024] In the above invention (Inventions 1 - 8), the injection means may inject the fluid toward the vicinity of the center of the outer surface of the rotary window (Invention 9). ]

[0025] According to such an invention (Invention 9), it is possible to more preferably remove water droplets or the like that remain near the center and cannot be completely removed.

[0026] In the above invention (Inventions 1 - 9), the driving means is a motor and a motor - side gear provided on the rotation shaft of the motor, the driven means is a frame - side gear provided across the outer periphery of the window frame of the rotary window, and the frame - side gear may be rotated by rotating the motor - side gear by the driving force of the motor (Invention 10).

Effect of the Invention

[0027] According to the estimation system of the present invention, clear image data near the object being cut using a cutting machine can be acquired regardless of the environment where a large amount of water, coolant, oil, etc. splashes, and the generation area of chips can be estimated based only on the acquired image data.

Brief Description of the Drawings

[0028] [Figure 1] This is a diagram showing an example of an estimation system according to one embodiment of the present invention. [Figure 2] This flowchart shows an example of a method for estimating the chip generation area using the estimation system shown in Figure 1. [Figure 3] This image shows an example (A) of the second still image Is2 that constitutes the second image data IDbf2, and an example (B) of the second extracted image Id2 created from this second still image Is2. [Figure 4] This is a configuration diagram showing an example of the hardware for the control circuit of the estimation device. [Figure 5] This is a configuration diagram showing an example of the hardware of the control circuit of an estimation device when it is implemented by software or firmware. [Figure 6] Figure 1 is a front view showing an example of an image acquisition device included in the estimation system. [Figure 7] Figure 6 is a cross-sectional view of the image acquisition device on line AA. [Figure 8] This is a partially enlarged cross-sectional view of the image acquisition device shown in Figure 6. [Modes for carrying out the invention]

[0029] Hereinafter, embodiments of the chip generation region estimation system of the present invention will be described with reference to the drawings.

[0030] [System for estimating the chip generation area] Figure 1 is a configuration diagram showing an example of a chip generation area estimation system (hereinafter simply referred to as the estimation system) according to one embodiment of the present invention. The estimation system 100 is a system that estimates the chip generation area generated by cutting an object using a cutting machine 200. The estimation system 100 comprises an image acquisition device 10 and an estimation device 20.

[0031] Figure 6 is a front view showing an example of the image acquisition device 10. Figure 7 is a cross-sectional view of the image acquisition device 10 along line AA in Figure 6. Note that in Figure 7, the up, down, left, and right directions are shown for clarity of explanation, but these directions do not represent the up, down, left, and right directions in actual use. The image acquisition device 10 includes a housing 13 having a hollow housing 11 and a rotating window 12 provided in an opening in the housing 11, a lens 14 positioned in the center of the housing 11 facing the rotating window 12, and a camera 15.

[0032] In the image acquisition device 10, the rotating window 12 has a light-transmitting window 12a, a window frame 12b that supports the window 12a at its periphery, and a driven means 16 provided on the outer circumference of the window frame 12b. The image acquisition device 10 is provided with a driving means 17 that rotates the rotating window 12 by driving the driven means 16 to rotate the window frame 12b, and a spraying means 18 that sprays fluid toward the outer surface of the window 12a. Not only does the rotating window 12 rotate, but the spraying means 18 also sprays fluid toward the outer surface of the window 12a, so that water splashes, oil stains, etc. that adhere to the rotating window 12 can be sufficiently removed. Therefore, clear image data IDs of the vicinity of the object being cut using the cutting machine 200 can be acquired regardless of the environment in which large amounts of water, coolant, oil, etc. are scattered.

[0033] The estimation device 20 estimates the chip generation area based on the image data IDs of the vicinity of the object acquired by the image acquisition device 10. According to the estimation system 100, the chip generation area can be estimated based only on the clear image data IDs acquired by the image acquisition device 10.

[0034] In this embodiment, the image data ID may include a moving image Im and a plurality of still images Is that constitute the moving image Im.

[0035] (estimation device) The estimation device 20 may estimate the chip generation area by specifying the chip generation area for the image data ID acquired by the image acquisition device 10 and monitoring the change in the area of ​​that area. As described above, if cutting is continued while the generated chips are left unattended, the workpiece may be damaged by the chips, affecting the accuracy of the workpiece. Furthermore, excessive chip generation can be an indicator of damage to the cutting edge of the cutting machine, errors in the setting of cutting parameters, etc. According to the estimation system 100, for example, if the area of ​​the chip generation area exceeds a predetermined value, it is possible to warn the user with alarm information. This allows for measures such as stopping the operation of the cutting machine 200 and removing the chips.

[0036] The estimation device 20 may include an extracted image creation unit 21 and a determination unit 22. The extracted image creation unit 21 creates multiple extracted image IDs, each specifying the chip generation area for each of the multiple still image Is that constitute the image data ID. The determination unit 22 determines whether the area of ​​the chip generation area exceeds a threshold for each of the multiple extracted image IDs. With this configuration, for example, when the determination unit 22 determines that the area of ​​the chip generation area for a particular extracted image ID exceeds a threshold, it is possible to alert the user with alarm information. The threshold can be set as appropriate as needed.

[0037] The extracted image creation unit 21 displays, for example, a closed boundary line indicating the chip generation area for each of the multiple extracted image Ids. That is, the extracted image creation unit 21 displays information regarding the position and size of the chip generation area for each of the multiple extracted image Ids. The determination unit 22 calculates, for example, the area inside the closed boundary line displayed for each of the multiple extracted image Ids and determines whether or not the area exceeds a threshold.

[0038] When the determination unit 22 detects an extracted image Id in which the area of ​​the chip generation region exceeds a threshold, it may transmit alarm information to an external display device 300 and / or the cutting machine 200. With such a configuration, for example, by checking the display device 300 which displays the alarm information, measures such as stopping the operation of the cutting machine 200 and removing the chips can be taken. For example, the operation of the cutting machine 200 can be controlled in advance so that it automatically stops when it receives the alarm information. Alternatively, the determination unit 22 may be configured to temporarily stop the operation of the cutting machine 200 when it receives the alarm information and to temporarily increase the injection pressure of the device that injects air or coolant inside the cutting machine 200 to remove the chips. In this case, the determination unit 22 may also be configured to automatically restart the operation of the cutting machine 200 when the area of ​​the chip generation region falls below a threshold.

[0039] The determination unit 22 may estimate the chip generation area in real time using a learning model that has been trained using training images with chips and non-training images without chips as training data from among multiple still images Is that constitute the image data ID. With such a configuration, the chip generation area can be estimated easily and accurately in real time.

[0040] The estimation system 100 may further include a storage device 30 having a storage unit 31 for temporarily storing image data IDs. When the determination unit 22 detects an extracted image Id in which the area of ​​the chip generation region exceeds a threshold, it checks the image data IDs before and after the threshold exceedance that are temporarily stored in the storage unit 31. bf The storage device 30 may also send an instruction to the extraction image creation unit 21 to send the image data ID. bf The extracted image may be transmitted to the image creation unit 21. In other words, the storage device 30 may have a skip-back function. With such a configuration, for example, the estimation accuracy of the chip generation area by the estimation device 20 can be improved.

[0041] In this embodiment, the image data ID includes the first image data ID1 acquired in advance by the image acquisition device 10 and the second image data ID2 acquired in real time by the image acquisition device 10. Therefore, the image data ID before and after exceeding the above threshold bf includes the image data ID before and after exceeding the above threshold bf which is the first image data ID acquired in advance by the image acquisition device 10 bf1 and the image data ID before and after exceeding the above threshold bf which is the second image data ID acquired in real time by the image acquisition device 10 bf2 is included. That is, not only the first image data ID bf1 is transmitted to the estimation device 20, but the second image data ID bf2 can also be transmitted.

[0042] The first image data ID1 (and the first image data ID bf1 ) is, for example, the image data ID acquired during the trial operation of the cutting machine 200. The second image data ID2 (and the second image data ID bf2 ) is, for example, the image data ID acquired during the normal operation of the cutting machine 200.

[0043] FIG. 2 is a flowchart showing an example of a method for estimating the chip generation area using the estimation system 100 of FIG. 1. For example, during the trial operation of the cutting machine 200, in the estimation device 20 that has received the first image data ID bf1 , the extracted image creation unit 21 uses the first image data ID bf1For each of the multiple first still images Is1 that make up the system, multiple first extracted images Id1 are created, each specifying the chip generation area. The determination unit 22 sorts the multiple first still images Is1 into a first training image Iy1 where chips are present and a first non-training image In1 where chips are not present. Furthermore, the determination unit 22 uses a learning model 25, which has been trained using the first training image Iy1, the multiple first extracted images Id1, and the first non-training image In1 as first training data, to estimate the chip generation area in real time. With this configuration, the chip generation area can be estimated simply and accurately in real time.

[0044] The learning model 25 is constructed, for example, by a neural network. As shown in Figure 2, the learning model 25 may also be constructed by deep learning. During training, the learning model 25 is given a first training image Iy1, a plurality of first extracted images Id1, and a first non-training image In1 as first training data. During training, when the first training data is given, the learning model 25 learns the chip generation area in the first training image Iy1. During inference, when the extracted image Id is given in real time, the learning model 25 outputs information indicating the chip generation area in the extracted image Id. In the estimation system 100 shown in Figure 1, the learning model 25 is located outside the estimation device 20. However, the learning model 25 may be located inside the estimation device 20.

[0045] For example, during normal operation of the cutting machine 200, the second image data ID bf2 Upon receiving the data, the estimation device 20 generates an extracted image ID, and the extraction image creation unit 21 generates the second image data ID. bf2For each of the multiple second still images Is2 that make up the image, multiple second extracted images Id2 are created, specifying the chip generation area. The determination unit 22 sorts the multiple second still images Is2 into second training images Iy2 in which chips are present and second non-training images In2 ​​in which chips are not present. Furthermore, the determination unit 22 estimates the chip generation area in real time using a learning model 25 that has been further trained using the second training images Iy2 as second training data. With this configuration, for example, the second training images Iy2 acquired during actual operation can be fed back, so the chip generation area can be estimated with greater accuracy in real time.

[0046] Figure 3 shows the second image data ID. bf2 This image shows an example of a second still image Is2 (A) that constitutes the second still image Is2, and an example of a second extracted image Id2 (B) created from this second still image Is2. As shown in Figure 3, the second extracted image Id2 is displayed as a heatmap image overlaid with an image in which the chip generation area identified in the second still image Is2 is colored. In Figure 3 (B), the area indicated by the arrow is the chip generation area displayed in heatmap format.

[0047] As shown in Figure 1, the image acquisition device 10, the estimation device 20, and the storage device 30 can each be connected via a network hub device 40. The network hub device 40 constitutes a network in a network system that includes a network such as Ethernet®.

[0048] Figure 4 is a configuration diagram showing an example of the hardware of the control circuit 23 of the estimation device 20. As described above, the estimation device 20 includes an extracted image creation unit 21 and a determination unit 22. The extracted image creation unit 21 is implemented, for example, by an extracted image creation circuit 23a. The determination unit 22 is implemented, for example, by a determination circuit 23b. The determination unit 22 uses a first training image Iy1, a plurality of first extracted images Id1, and a first non-training image In1 as first training data, and uses a second training image Iy2 as second training data.

[0049] Figure 5 is a configuration diagram showing an example of the hardware of the control circuit 24 of the estimation device 20 when it is implemented by software or firmware. When the estimation device 20 is implemented by software or firmware, a program that causes the computer to execute the respective processing procedures in the extracted image creation unit 21 and the determination unit 22 is stored in memory 24b. The computer's processor 24a executes the program stored in memory 24b.

[0050] Figure 4 shows an example where each component of the estimation device 20 is implemented by dedicated hardware. Figure 5 shows an example where the estimation device 20 is implemented by software or firmware, etc. However, for example, some components of the estimation device 20 may be implemented by dedicated hardware, and the remaining components may be implemented by software or firmware, etc.

[0051] (Image acquisition device) As described above, the image acquisition device 10 includes a housing 13, a lens 14, and a camera 15. The camera 15 comprises a camera body 15a, and the lens 14 is mounted on the camera body 15a. The camera body 15a and the lens 14 are fixed by screwing the lower end portion 15c of the lens 14, located on the lower side in Figure 7, into the inside of a fitting portion 15b that protrudes upward from the camera body 15a in Figure 7, and the lens 14 is removable. The camera body 15a is waterproof (and / or oilproof). On the opposite side of the lower end portion 15c of the lens 14, the leading lens element 15d of a plurality of lens elements (not shown) arranged inside the lens barrel is fixed.

[0052] The housing 13 has a hollow, roughly cylindrical housing 11, and a lens 14 is positioned inside the housing 11. The lower end portion 11a, which forms the lower opening in Figure 7 of the housing 11, is provided with a bottom portion 11b that closes the opening. The fitting portion 15b of the camera body 15a is fixed in the center of the bottom portion 11b, inserted into the housing 11. In addition, a supply means 11c for supplying gas (air) into the housing 11 is provided on the outer circumference of the camera body 15a at the bottom portion 11b. When air is supplied into the housing 11 by the supply means 11c, the inside of the housing 11 becomes positive pressure, which makes it possible to suppress the ingress of liquids into the housing 11 under the operating environment.

[0053] A rotating window 12 is provided at the upper end portion 11d of the housing 11, which forms the upper opening in Figure 7. The rotating window 12 comprises a transparent window 12a and a window frame 12b that holds the window 12a. In a front view, the window 12a is circular, and the window frame 12b is annular. The peripheral edge of the window 12a is fixed to the window frame 12b, and the circumferential end portion 12c of the window frame 12b engages with the upper end portion 11d of the housing 11. The window 12a and the lens element 15d of the lens 14 face each other, and the camera 15 can take pictures through the window 12a without its field of view being obstructed by any objects. The window frame 12b includes a cylindrical extension portion 12d that extends downward from the window 12a in Figure 7. A bearing 12e is provided on the outer circumference of the extension portion 12d, and the window frame 12b is fitted to the inner circumferential surface of the housing 11 via this bearing 12e. When fitting the inner circumferential surface of the housing 11 to the window frame 12b, in addition to the rolling bearing 12e, air bearings, oil-lubricated bearings, magnetic bearings, etc., may also be used.

[0054] The image acquisition device 10 further includes a driving means 17 for the rotating window 12. In this embodiment, the driving means 17 includes a motor 17a and a motor-side gear 17c provided on the rotation shaft 17b of the motor 17a. The rotating window 12, which rotates under the power of this driving force, is provided with a frame-side gear 16 (driven means) that is fitted to the motor-side gear 17c and is provided on the outer circumference of the extension portion 12d below the bearing 12e. The rotation speed of the rotating window 12 powered by the motor 17a is, for example, 3000 rpm. Note that the control wiring for the motor 17a is omitted.

[0055] In the image acquisition device 10 configured in this way, when the motor 17a is driven, the rotation of the rotation shaft 17b transmits the rotation of the motor-side gear 17c to the frame-side gear 16, thereby causing the rotating window 12 to rotate. When the rotating window 12 rotates, water droplets, oil droplets, dirt, etc. adhering to the outer surface of the window 12a are blown away, and the transparency of the outer surface of the window 12a can be maintained in good condition. However, in the vicinity of the rotation center 12f of the window 12a, centrifugal force cannot be obtained, so water droplets, oil droplets, etc. may exist in the form of a mist. Depending on the usage environment, it is possible that water droplets, oil droplets, dirt, etc. adhering to the window 12a cannot be blown away by rotation, and these may remain attached to the window 12a. Note that the vicinity of the rotation center 12f includes not only the center of the window 12a but also its surroundings.

[0056] Therefore, in this embodiment, the housing 11 is provided with an injection means 18 that injects gas toward the outer surface of the window 12a. In this embodiment, the injection means 18 is configured to inject air supplied to the housing 11 by a supply means 11c. Specifically, the injection means 18 comprises a communication pipe 18a that communicates with the inside of the housing 11 and a nozzle 18b provided on the communication pipe 18a, with the tip portion 18c of the nozzle 18b configured to face the vicinity 12f of the rotation center of the window 12a. In this embodiment, the inner diameter of the tip portion 18c of the nozzle 18b is formed to be smaller than that of the base portion of the nozzle 18b, and since air is supplied into the housing 11 from the supply means 11c and the inside of the housing 11 is under positive pressure, air is injected from the tip portion 18c of the nozzle 18b at an injection speed that can remove water droplets and the like adhering to the window 12a. The connecting pipe 18a may also be configured to have a compression means that compresses and discharges air, thereby forcefully ejecting air from the tip 18c of the nozzle 18b.

[0057] By having a spraying means 18, the image acquisition device 10 of this embodiment allows air supplied into the housing 11 by the supply means 11c to pass through the connecting pipe 18a and nozzle 18b, and to be sprayed from the nozzle at the tip 18c toward the center of the window 12a, thereby removing water droplets, oil droplets, and dirt adhering to the window 12a with air. In particular, in this embodiment, the shape and position of the tip 18c of the nozzle 18b are set so that the air released from the tip 18c hits the area near the center of rotation 12f of the window 12a. The centrifugal force is greater towards the periphery of the window 12a, making it easy for water droplets and oil droplets to scatter, but the centrifugal force is smaller near the center of rotation 12f, making it less likely for water droplets and oil droplets to scatter. Therefore, by having the air hit the area near the center of rotation 12f, it is possible to remove water droplets, oil droplets, dirt, etc. adhering to the window 12a more efficiently with air. Furthermore, in this embodiment, since the injection means 18 is configured to take out the air supplied to the housing 11 by the supply means 11c, the injection means 18 can be configured without adding any other supply means.

[0058] The injection by the injection means 18 may be performed, for example, at regular intervals (pulses), or it may be performed continuously. Furthermore, the fluid in this invention includes liquids and gases.In this embodiment, the shape of the tip portion 18c is configured to be narrower than the base side of the nozzle 18b, but it is not limited to this.For example, the tip portion 18c may have the same inner diameter as the nozzle 18b, or it may have a larger inner diameter than the nozzle 18b.In addition, the tip portion 18c may be configured to have multiple injection ports by branching into multiple parts, or a single tip portion 18c may have multiple injection ports.

[0059] Furthermore, in the image acquisition device 10 of this embodiment, the inside of the housing 11 is under positive pressure because air is supplied to the inside of the housing 11 by the supply means 11c, thereby suppressing the intrusion of water and other substances into the housing 11. In addition to this, as shown in Figure 8, the engagement portion 12g between the upper end portion 11d of the housing 11 and the circumferential end portion 12c of the window frame 12b is configured as a labyrinth structure. Specifically, the end of the upper end portion 11d is branched into multiple parts (two in this embodiment), and the circumferential end portion 12c also has multiple branches (two in this embodiment) formed along these branches, and these are fitted together to form a labyrinth structure in the engagement portion 12g. This makes it possible to further suppress the intrusion of water, oil, and other substances into the inside of the housing 11 without hindering the rotation of the rotating window 12.

[0060] As described above, air is supplied to the housing 11, so air is discharged from the labyrinth structure which is the engaging portion 12g. Here, since the terminal portion 12h at the circumferential end 12c of the window frame 12b is configured to cover the upper end portion 11d, the air is discharged downwards in Figure 8. Although the amount of air discharged from the engaging portion 12g is small, since it is near the window 12a, there is a concern that the discharge of air may stir up water or oil scattered in the environment and send it towards the window 12a. However, since the terminal portion 12h is configured to cover the upper end portion 11d, the air can be discharged in the opposite direction from the window 12a, thereby preventing water or oil scattered in the environment from being stirred up and sent towards the window 12a.

[0061] Furthermore, in the image acquisition device 10 of this embodiment, the housing 13 covers the entire lens 14, but the waterproof camera body 15a is positioned outside the housing 13. By not covering the waterproof camera body 15a with the housing 13, the entire device can be made compact, allowing it to be used in confined spaces. Additionally, the camera body 15a can be easily replaced, allowing the camera body 15a and lens 14 to be swapped depending on the application. By swapping to lenses with different focal lengths, the field of view can be changed according to the application. Moreover, the surface of the window 12a may be treated with a material having water-repellent / oil-repellent or hydrophilic / oil-lipophilic properties, or the window 12a itself may be constructed from such a material. In this case, the above-mentioned effects can be obtained even if the rotation speed of the rotating window 12 is low, which is advantageous in terms of extending lifespan, reducing noise, and reducing power consumption.

[0062] As the image acquisition device 10 of this embodiment, for example, the camera unit described in the international application PCT / JP2023 / 023254 (WO / 2023 / 249103) by the same applicant can be suitably used.

[0063] Typically, the cutting machine 200 is installed and operated inside a chamber to limit the environment from splashing large amounts of water, coolant, oil, etc. Although not shown in the figures, the image acquisition device 10 may further have a mounting part that allows it to be attached to the inner wall of the chamber of the cutting machine 200. With such a structure, the image acquisition device 10 can be mounted in an optimal position by the mounting part to more easily acquire image data IDs near the object.

[0064] One example of a mounting mechanism is one that uses magnets.

[0065] While embodiments of the present invention have been described above, the present invention is not limited to the embodiments described above. In the embodiments described above, an example was described in which the image acquisition device 10 has only one camera body 15a, but the image acquisition device 10 may have multiple camera bodies 15a. Also, in the embodiments described above, an example was described in which the estimation device 20 estimates the chip generation area using a learning model, but the estimation device 20 may estimate the chip generation area based on the image data ID acquired by the image acquisition device 10 without using a learning model.

[0066] The embodiments described above are provided to facilitate understanding of the present invention and are not intended to limit it. Accordingly, each element disclosed in the above embodiments is intended to include all design modifications and equivalents that fall within the technical scope of the present invention. [Explanation of Symbols]

[0067] 100 Estimation Systems 10 Image acquisition device 11 cabinets 12 Rotating windows 13 Housing 14 lenses 15 Cameras 16 Driven means 17 Driving means 18 Injection means 20 Estimation device 21 Extracted Image Creation Unit 22 Judgment section 23 Control circuits 23a Extraction Image Creation Circuit 23b Judgment circuit 24 Control circuits 24a processor 24b memory 25 Learning Models 30 Storage device 31 Storage section 40 Network Hub Devices 200 Cutting machine 300 display device

Claims

1. A system for estimating the chip generation area generated by cutting an object using a cutting machine, An image acquisition device including a housing having a hollow casing and a rotating window provided in an opening of the housing, a lens positioned in the center of the housing so as to face the rotating window, and a camera, An estimation device that estimates the chip generation area based on image data near the object acquired by the image acquisition device, Equipped with, The rotating window comprises a light-transmitting window, a window frame that supports the window at its peripheral edge, and a driven means provided on the outer circumference of the window frame. The chip generation region estimation system is provided with an image acquisition device which includes a driving means for rotating the rotating window by driving the driven means to rotate the window frame, and an injection means for injecting fluid toward the outer surface of the window.

2. The estimation system according to claim 1, wherein the estimation device includes an extracted image creation unit that creates a plurality of extracted images specifying the chip generation region for each of the plurality of still images constituting the image data acquired by the image acquisition device, and a determination unit that determines whether or not the area of ​​the chip generation region exceeds a threshold for each of the plurality of extracted images.

3. The estimation system according to claim 2, wherein the determination unit outputs alarm information to an external display device or the cutting machine when it detects in the extracted image that the area of ​​the chip generation region exceeds a threshold.

4. The storage device further comprises a storage unit for temporarily storing the aforementioned image data, The estimation system according to claim 2, wherein when the determination unit detects an extracted image in which the area of ​​the chip generation region exceeds a threshold, the storage device transmits the image data before and after the threshold exceedance, which is temporarily stored in the storage unit, to the extracted image creation unit.

5. The image data before and after exceeding the threshold includes a first image data acquired in advance by the image acquisition device and a second image data acquired in real time by the image acquisition device. The extracted image creation unit creates a plurality of first extracted images for each of the plurality of first still images constituting the first image data, specifying the chip generation region. The estimation system according to claim 4, wherein the determination unit sorts the plurality of first still images into first training images in which chips are present and first non-training images in which chips are not present, and estimates the chip generation area using a learning model trained with the first training images, the plurality of first extracted images, and the first non-training images as first training data.

6. The extracted image creation unit creates a plurality of second extracted images for each of the plurality of second still images that constitute the second image data, specifying the chip generation region. The estimation system according to claim 5, wherein the determination unit sorts the plurality of second still images into second training images in which chips are present and second non-training images in which chips are not present, and estimates the chip generation area using the learning model which has been further trained using the second training images as second training data.

7. The aforementioned cutting machine is installed inside the chamber. The estimation system according to claim 1, wherein the image acquisition device is provided with a mounting portion that allows the image acquisition device to be attached to the inner wall of the chamber.

8. The housing further includes a supply means for supplying gas into the housing, The estimation system according to claim 1, wherein the injection means is configured to inject the gas supplied by the supply means toward the outer surface of the window.

9. The estimation system according to claim 1, wherein the injection means injects the fluid toward the vicinity of the center of the outer surface of the rotating window.

10. The driving means is a motor and a motor-side gear provided on the rotating shaft of the motor, and the driven means is a frame-side gear provided over the outer circumference of the window frame of the rotating window. The estimation system according to claim 1, wherein the motor-side gear is rotated by the driving force of the motor, thereby rotating the frame-side gear.

Images