Tool magazine tool detection device and method

Abstract

The application discloses a tool magazine tool detection device and method, which comprises a fixed rack, a tool self-rotation module, a air-drying module, a lateral visual detection module, an axial visual detection module and a control system. The tool self-rotation module is arranged at a fixed tool changing point of a real tool magazine, and is fixedly connected with the axial visual detection module perpendicularly on the fixed rack in the device, and the lateral visual detection module and the air-drying module are fixedly connected on the axial visual detection module. The tool self-rotation module is arranged at a tool changing position of a combined tool magazine, and is composed of a rotating mechanism and a tool handle clamp mounted on the rotating mechanism, so that the tool self-rotation function is realized when the tool handle is fixed and detected. The lateral visual detection module and the axial visual detection module are both composed of a visual detection device and a driving device, so that the tool damage surface picture collection function is realized. The air-drying module blows the gas from an external source to the tool after the gas is discharged through a slit, so that the tool cleaning and air-drying function is realized. The application can collect the bottom and side edge picture wear information of multiple types of tools in the tool magazine through vision.

Description

Technical Field

[0001] This invention belongs to the field of mechanical automation and image processing technology, specifically relating to a tool detection device and method integrated on a tool magazine. Background Technology

[0002] With the continuous advancement of science and technology, the machining and manufacturing field is gradually developing towards higher precision and higher efficiency. Multifunctional composite machining centers have become a major development direction for CNC equipment. To pursue higher machining efficiency, large-scale combined tool magazines are often used as auxiliary equipment. During machining, tool wear directly affects the machining accuracy of the product. Therefore, how to detect the degree of tool wear, determine the tool's health status, and replace the tool in a timely manner during tool changing is of great significance for avoiding excessive tool wear that leads to reduced machining quality and other more serious losses, improving machining efficiency, and reducing machining costs.

[0003] Visual inspection is widely used in the manufacturing industry, but its application in inspecting multiple models and types of cutting tools in tool magazines is currently lacking. Furthermore, most existing visual inspection technologies for cutting tools are still in the offline inspection stage, requiring manual intervention to perform static photographic inspections. For complex tool shapes and edge wear inspections, manual repositioning is necessary, potentially requiring multiple clamping operations, resulting in low efficiency. Additionally, re-calibration of focus is required when inspecting tools of different specifications, leading to inconsistent inspection accuracy.

[0004] The current tool wear index is the standard specified by ISO for testing the durability of lathe tools. When the cutting edge wear band is uniform, it is considered dull when the average wear band VB > 0.3 mm; when the cutting edge wear band is uneven, the maximum wear band VB is considered dull.

[0005] VB max A wear level greater than 0.6 mm is considered dull. Due to differences in cutting edge shapes, machining forces, and tool durability among different tools, the resulting wear zone morphology and dullness standards vary. Using a single dullness standard to characterize the health of all tools can lead to inaccurate detection. Furthermore, current standards blur the concepts of wear and breakage, uniformly defining both as wear areas and assuming they have the same impact on tool life, when in reality, breakage areas have a significantly greater influence. Therefore, establishing a multi-dimensional tool dullness evaluation method based on image processing results and expert scoring to determine the weights of wear and breakage in affecting tool life is a valuable supplement to existing dullness standards. Summary of the Invention

[0006] The purpose of this invention is to provide a tool detection device and corresponding detection method on an integrated tool magazine, which can efficiently detect the wear of various types and models of tools in the tool magazine through vision, and realize the tool health detection of composite machining centers.

[0007] The technical solution to achieve the purpose of this invention is as follows:

[0008] A tool magazine tool detection device, comprising:

[0009] A fixed stand is used to install the spin module, drying module, lateral vision inspection module, and axial vision inspection module, and is set at the fixed tool change point of the lower tool magazine;

[0010] The spin module is used to fix the tool holder and rotate the tool during inspection to facilitate the acquisition of side and bottom edge images at different angles.

[0011] The lateral vision inspection module is used to acquire images of the side edge of the tool and can move in the horizontal direction perpendicular to the rotation axis of the spin module, and is equipped with a first detection unit.

[0012] The axial vision inspection module is used to acquire images of the bottom edge of the tool and can move in a direction parallel to the rotation axis of the spin module.

[0013] The air-drying module, located between the spin module and the axial vision inspection module, is used to spray airflow toward the tool direction to remove cutting fluid and debris adhering to the tool to be inspected. It can move together with the axial vision inspection module and has a second inspection unit inside to control the movement position of the axial vision inspection module and the movement of the lateral vision inspection module.

[0014] The control system controls the spin module, drying module, lateral vision inspection module, and axial vision inspection module. During inspection, the spin module rotates, the axial vision inspection module moves toward the tool, and the drying module is activated. When the second inspection unit detects the tool, the axial vision inspection module stops moving, ensuring that the bottom edge of the tool is within the imaging depth of the axial vision inspection module. Simultaneously, the lateral vision inspection module begins moving. When the first inspection unit detects the tool, the lateral vision inspection module stops moving, ensuring that the side edge of the tool is within the imaging depth of the lateral vision inspection module.

[0015] A detection method for a tool magazine tool detection device includes:

[0016] Image acquisition of the bottom and side edges of the cutting tool;

[0017] Image preprocessing: Obtain a circular image containing all bottom edge wear information and enhance the contrast between the worn and non-worn areas;

[0018] Preliminary acquisition of the tool wear area;

[0019] Preliminary acquisition of tool wear lines;

[0020] Precise acquisition of tool wear area;

[0021] Template matching-based acquisition of broken bottom edge regions: Template matching is performed between images of new and old bottom edges stored in the database. Based on the template matching, the association information of non-wear areas of the new and old blades is identified, thereby obtaining the angle θ that the template image needs to be rotated when the two images completely overlap. After preprocessing the new blade image, threshold segmentation is directly performed to obtain a binary image of the complete bottom edge before wear. This image is used as a template and rotated by an angle θ. It is then compared with the binary image of the precise wear area of ​​the bottom edge. The resulting image is a binary image of the bottom edge of the new blade with missing wear areas. Finally, area filtering is used to retain the area with the smallest outline area, which is the corresponding broken bottom edge region.

[0022] Side blade damage area acquisition based on blade line reshaping: Detect and identify the straight line containing the unworn blade line in the old knife's side blade image. Reshape the blade line according to the obtained straight line coordinate parameters. Perform contour re-identification on the reshaped image and use it as the image of the new knife. After preprocessing the reshaped new knife image, perform threshold segmentation to obtain a binarized image. Perform difference processing with the binary image of the accurate wear area of ​​the side blade. The obtained image is the binary image of the new knife's side blade with missing wear areas. Then, through area filtering, retain the area with the smallest contour area, which is the corresponding side blade damage area.

[0023] Collect basic parameters of the grinding and damage profiles, including the area and bandwidth of the grinding and damage profile regions;

[0024] Establish and improve tool life evaluation indicators, including three-dimensional standards: maximum bandwidth scrapping standard, average bandwidth scrapping standard, two-dimensional maximum area scrapping standard, and three-dimensional volume scrapping standard.

[0025] The significant advantages of this invention compared to existing technologies are:

[0026] (1) In this invention, with the cooperation of the tool spin module, axial and lateral vision inspection modules, the bottom edge detection camera takes pictures to identify the angle between each edge. Utilizing the characteristics of tool spin, the identified angle is fed back to the servo motor of the tool spin module, enabling the side edge detection camera to comprehensively collect information on all side edges. This achieves all-round identification of the wear condition of the side and bottom edges of the tool to be inspected within two mutually perpendicular translation paths.

[0027] (2) The present invention can inspect the replacement tools on the production line before they are put into storage. There is no need to refocus between the inspection and conversion process of different types of tools, and there is no need to clamp the same tool multiple times during the inspection process, which greatly improves the efficiency and quality of inspection.

[0028] (3) The wear area was separated by combining the damage morphology of the bottom edge and the side edge. Based on the stable characteristics of the damage images collected by the invented device, a method for obtaining the bottom edge damage area based on template matching rotational difference and a method for obtaining the side edge damage area based on the edge line reshaping matching difference were proposed. The wear and damage areas were distinguished, and the degree of tool damage was comprehensively evaluated based on the damage conditions of the two parts.

[0029] (4) Based on expert scoring, the weight of the impact of grinding and broken areas on tool life is given, and a multi-evaluation system for tool damage is established based on the weight. With this evaluation system, the health status of tools in the tool magazine can be more accurately grasped. Attached Figure Description

[0030] Figure 1 This is a schematic diagram of the structure of an embodiment of the present invention.

[0031] Figure 2 This is a cross-sectional view of the tool spin module.

[0032] Figure 3 This is a detailed image of the air-drying module.

[0033] Figure 4 This is a detailed diagram of the lateral vision inspection module.

[0034] Figure 5 This is a detailed diagram of the axial vision inspection module.

[0035] Figure 6 Flowchart for image acquisition by the tool inspection device.

[0036] Figure 7 A step-by-step diagram is obtained for the grinding and damaged areas of the cutting tool.

[0037] Figure 8 This is a schematic diagram of damage to the bottom and side edges of a cutting tool.

[0038] Figure 9 This is a schematic diagram of obtaining the broken area of ​​the bottom edge based on template matching.

[0039] Figure 10 A schematic diagram of the side edge damage area based on the reshaping of the cutting edge line.

[0040] Figure 11 This is a logic diagram for evaluating tool life. Detailed Implementation

[0041] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0042] Combination Figure 1This embodiment of a tool magazine tool detection device includes a fixed stand 1, a spin module 2, a drying module 3, a lateral vision detection module 4, an axial vision detection module 5, and a control system 6. The fixed stand 1 is constructed from aluminum profiles and is installed at a fixed tool change point in a real-world tool magazine environment. The spin module 2 and the axial vision detection module 5 are both fixed to the fixed stand 1. The lateral vision detection module 4 and the drying module 3 are both fixed to the axial vision detection module 5 according to a preset tool detection distance. The rotation axis of the spin module 2 is parallel to the movement direction of the axial vision detection module 5 and perpendicular to the movement direction of the lateral vision detection module 4. Simultaneously, the rotation axis of the spin module 2 is at the same height as the axial vision module 5 and the camera lens axis of the lateral vision detection module.

[0043] Combination Figure 2 The spin module 2 includes a servo motor 201, a reducer 202, a motor base rear cover 203, a locking nut 204, a bearing 205, a tool holder connector 206, a motor base front cover 207, a rotation module support plate 208, a spring plunger 209, a tool holder 2010, and a tool holder 2011 containing a tool. The spin module is driven by the servo motor 201 connected to the reducer 202. The output end of the servo motor 201 is connected to the input end of the reducer 202 and fixed to the power supply. On the rear cover 203 of the machine base, one end of the tool holder connector 206 is connected to the output end of the reducer 202 via a keyway, and the other end is fixed to the tool holder 2010 via a flange. The tool holder connector 206 is rotatably connected to the front cover 207 of the motor base via a bearing 205 and a locking nut 204. The rear cover 203 of the motor base is connected to the front cover 207 of the motor base. The front cover 207 of the motor base is fixed to the rotating module support plate 208, and the rotating module support plate 208 is fixed to the fixed frame 1. In addition, the tool holder is inserted into the tool holder 2010 by the tool magazine changer. Four spring plungers 209 are evenly distributed at the end of the tool holder 2010 to clamp the tool holder pull studs and fix the tool holder 2011 by tension.

[0044] Combination Figure 3 The air-drying module 3 includes an axial tool position detection through-beam sensor 301, a high-speed airflow cavity 302, a vent connector 303, and a fixed bracket fixture 304. The high-speed airflow cavity 302 is an annular structure with an internal airflow cavity. Multiple slits are formed on its inner surface along the feed direction of the tool to be tested, and a threaded hole is formed on its outer surface to connect to the vent connector 303. The cavity, slits, and vent connector are interconnected. Threaded holes are formed on the upper and lower surfaces of the high-speed airflow cavity 302, and the axial tool position detection through-beam sensor 301 is symmetrically installed thereon. The right side is fixed to the fixed bracket fixture 304 via an adjustable sliding clamp. The fixed bracket fixture 304 is positioned vertically (…). Figure 1The position of the drying module 3 in the Z-direction is adjustable. After the external gas is connected to the air connector 303, it is compressed through the cavity slit and sprayed as a high-speed airflow towards the tool.

[0045] Combination Figure 4 The lateral vision detection module 4 includes a first moving unit 401, a first slide block 402, a diffuse reflection sensor and its fixture 403 for the lateral vision module, a first fine-tuning slide 404, a side-blade detection light source 405, a side-blade detection lens 406, and a side-blade detection camera 407. The first moving unit 401 has one degree of freedom of movement in one direction. Figure 1 In the Y-direction (the horizontal direction perpendicular to the movement direction of the second moving unit 501), the first slide block 402 is fixed to the first moving unit 401 and can move with the first moving unit 401 (motor, lead screw mechanism). The diffuse reflection sensor and its fixture 403 consist of a fixture bracket and a diffuse reflection sensor fixed on it by threads. Finally, according to the preset detection distance of the tool side edge, it is arranged in a suitable position and fixed to the front end of the first slide block 402 by screws. The first fine-tuning slide 404 is fixed to the rear end of the first slide block 402. On it, the side edge detection light source 405, the side edge detection lens 406, and the side edge detection camera 407 are respectively arranged from front to back by screws according to the working parameters of the vision inspection device.

[0046] Combination Figure 5 The axial vision detection module 5 includes a second moving unit 501, a second slide block 502, a second fine-tuning slide 503, a bottom edge detection light source 504, a bottom edge detection lens 505, a bottom edge detection camera 506, and a support plate 507. The second moving unit 501 is fixed to the support plate 507 and has one degree of freedom of movement in one direction. Figure 1 In the X-direction (the direction of movement is parallel to the tool axis), the second slide 502 is fixedly connected to the second moving unit 501 (motor and lead screw mechanism) and can move with the second moving unit 501. The second fine-tuning slide 503 is fixed to the rear end of the second slide 502, and on it, according to the working parameters of the vision inspection device, the bottom edge detection light source 504, the bottom edge detection lens 505, and the bottom edge detection camera 506 are respectively arranged from front to back via screws through sheet metal parts. The support plate 507 is fixedly connected to the fixed frame 1.

[0047] Combination Figure 2 The tool holder connector 206 of the spin module 2 can be connected to different types of tool holders or other workpieces to be tested, adapting to the tool testing needs in different tool magazines.

[0048] Combination Figure 4 , 5In the lateral vision inspection module 4, the diffuse reflection sensor and its fixture 403 are arranged directly in front of the side blade detection light source 405, and the arrangement distance is determined according to the working parameters of the vision device. The lateral vision inspection module 4 is reasonably arranged at the installation position of the second slide 502 according to the lens parameters of the bottom and side blade detection cameras.

[0049] Combination Figure 4 , 5 The axial vision detection module 5 and the lateral vision detection module 4 use the same coaxial positive white light source, fixed-focus low-distortion ordinary lens, and global black and white industrial camera to acquire images of the tool bottom and side edge.

[0050] Combination Figure 4 , 5 The first fine-tuning slide 404 in the lateral vision detection module 4 provides small-range movement of the side blade detection vision device in the X and Z axis directions, and the second fine-tuning slide 503 in the axial vision detection module 5 provides small-range movement of the bottom blade detection vision device in the Y and Z axis directions, thereby expanding the detection types.

[0051] Combination Figure 2 , 3 4, 5, the axis of the drying module 3, bottom and side blade image acquisition device is adjustable in height in the Z-axis direction, and needs to be adjusted to the same height as the axis of the spin module 2 before detection.

[0052] Combination Figures 6 to 10 This example describes a tool life detection method using the aforementioned tool detection device. The detection process and method are as follows:

[0053] Step 1: Adjust the camera lens light source:

[0054] Adjust the light source brightness, lens aperture, lens focal length, and relative distances between the lens and the light source, and between the light source and the sensor at the camera detection position until the best tool imaging image can be obtained. Then fix all parameters and relative positions. Finally, calibrate the camera at the detection position to facilitate the subsequent conversion of pixel parameters of the recognition area into physical parameters.

[0055] Step 2: Image acquisition of the tool bottom and side cutting edge:

[0056] (1) After the tool magazine executes the tool change command, the mechanical tool changer inserts the tool used by the machining center into the tool holder of the tool spin module and starts to start the device detection status. The bottom edge image acquisition device in the axial vision detection module and the air drying module and the side vision detection module on it move towards the tool to be detected under the drive of the second motor.

[0057] (2) The motor in the tool spin module starts and drives the tool to spin. At the same time, the air supply of the air drying and cleaning device is turned on, blowing a high-speed airflow in the direction of the tool to be tested.

[0058] (3) The cutter passes through the air-drying and cleaning device, blocking the photoelectric signal of the through-beam sensor arranged on it. The axial vision detection module of the control system stops the motor, and at this time, the bottom edge of the cutter is within the imaging depth of the bottom edge detection camera. Subsequently, the lateral vision detection module drives the side edge image acquisition device and the diffuse reflection sensor installed at its front end to move towards the cutter to be inspected. When the cutter body blocks the photoelectric signal of the diffuse reflection sensor, the motor in the lateral vision detection module stops running, and at this time, the side edge of the cutter is within the imaging depth of the side edge detection camera. Then, the air supply of the air-drying module is turned off.

[0059] (4) The light source in the bottom edge detection image acquisition device is turned on, the camera takes a picture of the bottom edge of the tool and uploads it, and then the light source is turned off. Then the light source in the side edge image acquisition device is turned on. During the tool spin process, the side edge detection camera collects a set of pictures containing comprehensive wear information of each edge surface at a certain frame rate and uploads them to the host computer, and then the light source is turned off.

[0060] (5) After the edge surface image acquisition is completed, the lateral and axial vision detection modules return to their origin under the drive of their respective drive motors; the spin module stops spinning.

[0061] In particular, the blade image of each new knife will be captured before it is put into storage, and stored as a blade wear comparison template.

[0062] Step 3: Image processing to separate the grinding and damaged areas of the tool and collect important parameters:

[0063] (1) Image preprocessing: First, identify the center point of the specific imaging part of the blade in the bottom edge image, and take this point as the center to crop the largest circular field of view in the imaging field of view to obtain a circular image containing all the bottom edge wear information, which is convenient for subsequent bottom edge rotation matching; then, perform conventional preprocessing on the processed circular bottom edge image and the obtained side edge image, and use bilateral filter to filter the obtained image to remove noise while maintaining the clarity and sharpness of the wear edge line; finally, enhance the contrast between the wear area and the non-wear area by changing the gray index, which is convenient for subsequent separation of the wear area and the wear edge line.

[0064] (2) Preliminary acquisition of tool wear area: The preprocessed image is divided into thresholds, the gray-level histogram of the image is analyzed, and the gray value corresponding to the valley of the gray-level histogram is found according to the maximum variance between classes. This gray value is used as the threshold division basis to separate the wear area and place it as the foreground. Then, morphological composite processing is used to extract the backbone information of the wear area, weaken the cutting lines that are not closely connected to the backbone area, and eliminate isolated wear points.

[0065] (3) Preliminary acquisition of tool wear edge line: The edge operator is used to traverse the preprocessed image, calculate the gradient magnitude and direction of the wear edge, perform non-maximum suppression on the gradient magnitude to retain strong edges and remove weak edges, and adjust the threshold to obtain the ideal edge connection state.

[0066] (4) Precise acquisition of tool wear area: The above methods for wear area separation and wear edge line separation have different principles and their processing results have their own advantages and disadvantages. Now, the processing results of the two are superimposed for edge enhancement, and the boundary identified by both is taken as the final wear boundary. Contour search is performed on the image after edge enhancement. The required wear area at the tool tip is selected according to the contour area and a mask is built accordingly. The gray level of the pixels inside the selected contour is set to 255 as the foreground and the gray level of the pixels outside the contour is set to 0 as the background, thus obtaining a binary image containing only the target wear area.

[0067] (5) Obtaining the broken area of ​​the bottom edge based on template matching: Template matching is performed between the images of new and old bottom edges stored in the database. Based on the template matching, the association information of the non-wear areas of the new and old blades is identified, thereby obtaining the angle θ that the template image needs to be rotated when the two images completely overlap. After preprocessing the new blade image, threshold segmentation is directly performed to obtain a binary image of the complete bottom edge before wear. This image is used as the template after rotating by an angle θ and is then compared with the binary image of the accurate wear area of ​​the bottom edge processed in (1) to (4). The resulting image is a binary image of the bottom edge of the new blade with missing wear area. The area with the smallest contour area is then retained by area filtering, which is the corresponding broken area of ​​the bottom edge. See details. Figure 9 .

[0068] (6) Acquisition of the damaged side blade area based on blade line reshaping: The straight line containing the unworn blade line in the old knife's side blade image is identified by Hough line detection. The blade line is reshaped according to the obtained line coordinate parameters. The reshaped image is then re-identified based on the contour and used as the image of the new knife. After preprocessing the reshaped new knife image, threshold segmentation is directly performed to obtain a binarized image. This binarized image is then compared with the binary image of the accurate wear area of ​​the side blade processed in (1) to (4). The resulting image is a binary image of the new knife's side blade with missing wear area. The area with the smallest contour area is then selected by area filtering, which is the corresponding damaged side blade area. See details. Figure 10 .

[0069] (7) Collect basic parameters of the wear and damage contour: Enclose the wear and damage contour with a minimum bounding rectangle. The width of this rectangle is the maximum wear bandwidth of the wear and damage area. Identify the area of ​​the wear and damage contour area, and construct a rectangle of equal area using the area of ​​the wear and damage area and the length of the minimum bounding rectangle. The width of this rectangle is the average wear bandwidth of the wear and damage area. The obtained bandwidth, area, and other parameters should be converted into physical values ​​in conjunction with the camera pixel calibration results.

[0070] Step 4: Establish and improve tool life evaluation indicators:

[0071] (1) Based on expert scoring, determine the weights A and B of the impact of tool wear and breakage on tool life, and establish a one-dimensional maximum bandwidth scrapping standard. Average bandwidth obsolescence standard (VB) max Two-dimensional maximum area scrapping standard S max and three-dimensional volume scrapping standard V max .

[0072] (2) Combination Figure 11 Referring to current cutting tool dulling standards, when the flank wear bandwidth is uniform, the average flank wear width is used as the dulling standard; when the flank wear bandwidth is uneven, the maximum flank wear width is used as the dulling standard. A one-dimensional dulling standard is: if the bottom side grinding and damaged area is a uniform band, it meets the following criteria. Consider it as a dull tool; if the bottom edge is worn or the damaged area is an uneven band, then it satisfies VB. *m ×A+vb *p ×B>VB max This is considered as the tool becoming dull. The average bandwidth of the wear area is represented by "*" (d represents the bottom edge; "*" represents the side edge, and so on). Represents the average bandwidth of the damaged area; VB *m and VB *p These represent the maximum bandwidth of the wear and breakage areas, respectively; A and B are the set weights for the impact of lifespan. and VB max Both are established as one-dimensional scrapping standards, representing the maximum average damage bandwidth and the maximum damage bandwidth, respectively.

[0073] Two-dimensional wear standard: when the wear area s *m Damaged area s *p Satisfy s *m ×A+s *p ×B>S max When this happens, it is considered that the tool is dull;

[0074] Three-dimensional wear standard: The equivalent damage area s obtained by combining the bottom edge grinding and the area of ​​the damaged area with influence weights. dm ×A+s dp ×B is considered as the base area corresponding to the volume of the broken blade tip. The length of the damage band corresponding to the equivalent damage area is then obtained by combining the side edge grinding and the area of ​​the broken area with the influence weights. Multiply by the different blade shape ratio coefficients λ to obtain the corresponding height of the blade tip breakage volume; when the fitted volume satisfies When the tool reaches a dull state, it is considered dull. If any one of the above three criteria is met, the tool is determined to be dull, and an alarm signal will be sent to the machining center to prompt the replacement of the tool.

[0075] The implementation process is as follows:

[0076] In the non-detection state, the first moving unit 401 in the lateral vision detection module and the second moving unit 501 in the axial vision detection module are located at the origin, i.e., away from the tool spin module end. When changing tools between the composite machining center and the combined tool magazine, the mechanical tool changer inserts the tool used by the machining center into the tool sleeve 2010 of the tool spin module. When the detection program needs to be activated, the control system sends a command to the detection device to complete the detection process. First, driven by the second moving unit 501 of the axial vision detection module, the second slide 502, along with other modules connected to it, moves towards the tool to be detected. As the tool to be detected approaches, the drying module 3 is activated, and an external air source rushes into the high-speed airflow cavity 302 and sprays a high-speed airflow towards the tool feed end through a slit, removing the cutting fluid and debris adhering to the tool to be detected, making it easier for the camera to obtain a clearer picture of the tool wear. Simultaneously, sensor 301 in the drying module detects the bottom edge of the tool entering the field of view of bottom edge detection camera 506. Once it enters, the second moving unit stops. At the same time, the first moving unit 401 of the lateral vision detection module starts to drive, causing the first slide 402 and the side edge detection module on it to approach the tool to be detected. The sensor placed in front of the side edge detection light source 405 is used to detect the side edge of the tool entering the field of view of side edge detection camera 407. Once it enters, the first moving unit stops. At this time, the tool wear line is at the intersection of the fields of view of the two camera modules. Next, two industrial cameras (side edge detection camera 407 and bottom edge detection camera 506) take pictures of the bottom and side edges of the tool under test, respectively. Since most tools have multiple cutting edges, and the included angles between the cutting edges are relatively fixed, the bottom edge detection camera 506 identifies the included angles between each cutting edge after taking the bottom edge picture, and then feeds the included angle information back to the control system. This controls the servo motor 201 in the spin module to rotate, driving the tool to rotate by the corresponding angle, so that the side edge detection camera 407 can take pictures and identify all the cutting edges until all cutting edges are identified. After the tool under test completes the entire inspection, the first moving unit 401 and the second moving unit 501 reverse drive, returning the lateral vision inspection module 4 and the axial vision inspection module 5 to their respective origins until the next inspection task arrives. When occasionally a few tool end faces under inspection are too large, the first fine-tuning slide 404 and the second fine-tuning slide 503 can be adjusted to expand the inspection range.

[0077] The above description is merely an embodiment of the present invention. In practice, the invention also considers the detection of other common wear-prone workpieces. Detection can be achieved simply by connecting the parts to the flange connector and ensuring the workpiece can pass smoothly through the drying device. Other well-known structural features and characteristics are not described in detail here. However, it should be emphasized that modifications and improvements made based on the structure of this invention should also be considered within the scope of protection of this invention for those skilled in the art.

Claims

1. A method for detecting cutting tools in a tool magazine, characterized in that, include: Image acquisition of the bottom and side edges of the cutting tool; Image preprocessing: Obtain a circular image containing all bottom edge wear information and enhance the contrast between the worn and non-worn areas; Preliminary acquisition of the tool wear area; Preliminary acquisition of tool wear lines; Precise acquisition of tool wear area; Template matching-based acquisition of broken bottom edge regions: Template matching is performed between images of new and old bottom edges stored in the database. Based on the template matching, the association information of non-wear areas of the new and old blades is identified, thereby obtaining the angle θ that the template image needs to be rotated when the two images completely overlap. After preprocessing the new blade image, threshold segmentation is directly performed to obtain a binary image of the complete bottom edge before wear. This image is used as a template and rotated by an angle θ. It is then compared with the binary image of the precise wear area of ​​the bottom edge. The resulting image is a binary image of the bottom edge of the new blade with missing wear areas. Finally, area filtering is used to retain the area with the smallest outline area, which is the corresponding broken bottom edge region. Side blade damage area acquisition based on blade line reshaping: Detect and identify the straight line containing the unworn blade line in the old knife's side blade image. Reshape the blade line according to the obtained straight line coordinate parameters. Perform contour re-identification on the reshaped image and use it as the image of the new knife. After preprocessing the reshaped new knife image, perform threshold segmentation to obtain a binarized image. Perform difference processing with the binary image of the accurate wear area of ​​the side blade. The obtained image is the binary image of the new knife's side blade with missing wear areas. Then, through area filtering, retain the area with the smallest contour area, which is the corresponding side blade damage area. Collect basic parameters of the grinding and damage profiles, including the area and bandwidth of the grinding and damage profile regions; Establish and improve tool life evaluation indicators, including three-dimensional standards: maximum bandwidth scrapping standard, average bandwidth scrapping standard, two-dimensional maximum area scrapping standard, and three-dimensional volume scrapping standard.

2. The detection method according to claim 1, characterized in that, The one-dimensional dulling standard is: if the bottom edge is ground and the damaged area is a uniform band, it meets the following criteria. Consider it as a dull tool; If the bottom edge grinding and damaged area is an uneven band, then VB is satisfied. *m ×A+vb *p ×B>VB max Considered as a dull tool; among which This represents the average bandwidth of the wear area. Represents the average bandwidth of the damaged area, VB *m and VB *p These represent the maximum bandwidth of the wear and breakage areas, respectively, with A and B being the set weights for the impact of lifespan. and VB max Both are established as one-dimensional scrapping standards, representing the maximum average damage bandwidth and the maximum damage bandwidth, respectively. Two-dimensional wear standard: when the wear area s *m Damaged area s *p Satisfy s *m ×A+s *p ×B>S max When this happens, it is considered that the tool is dull; In the superscript and subscript, "*" represents d or c, which represent the bottom edge or the side edge, respectively. Three-dimensional wear standard: The equivalent damage area s obtained by combining the bottom edge grinding and the area of ​​the damaged area with influence weights. dm ×A+s dp ×B is considered as the base area corresponding to the volume of the broken blade tip. The length of the damage band corresponding to the equivalent damage area is then obtained by combining the side edge grinding and the area of ​​the broken area with the influence weights. Multiply by the different blade shape ratio coefficients λ to obtain the corresponding height of the blade tip breakage volume; when the fitted volume satisfies At this time, observe whether the cutting tool has reached a dull state; If any one or more of the above three criteria are met, the tool is considered dull.

3. The detection method according to claim 1, characterized in that, The process of collecting basic parameters of the wear and breakage profile is as follows: the wear and breakage profile is enclosed by a rectangle with the minimum bounding rectangle. The width of this rectangle is the maximum wear bandwidth of the wear and breakage area. The area of ​​the wear and breakage profile area is identified, and a rectangle with equal area is constructed using the area of ​​the wear and breakage area and the length of the minimum bounding rectangle. The width of this rectangle is the average wear bandwidth of the wear and breakage area.

4. The detection method according to claim 1, characterized in that, The image preprocessing process involves first identifying the center point of the specific imaging portion of the blade in the bottom edge image, and then cropping the largest circular field of view outward from this point to obtain a circular image containing all the bottom edge wear information. Next, the processed circular bottom edge image and the obtained side edge image are subjected to conventional preprocessing, using a bilateral filter to remove noise while maintaining the clarity and sharpness of the wear edge edges. Finally, the contrast between the worn and non-worn areas is enhanced by changing the grayscale index.

5. The detection method according to claim 1, characterized in that, The initial process for obtaining the tool wear area is as follows: the preprocessed image is thresholded, the gray-level histogram of the image is analyzed, the gray-level value corresponding to the valley of the gray-level histogram is found according to the maximum variance between classes, and this gray-level value is used as the threshold for segmentation. The wear area is separated and placed as the foreground; the backbone information of the wear area is extracted, the cutting lines that are not closely connected to the backbone area are weakened, and isolated wear points are eliminated. The initial process for obtaining the tool wear edge line is as follows: using an edge operator to traverse the preprocessed image, calculate the gradient magnitude and direction of the wear edge, perform non-maximum suppression on the gradient magnitude to retain strong edges and remove weak edges, and adjust the threshold to obtain the edge connectivity status.

6. A tool magazine detection device for use in the detection method according to any one of claims 1-5, characterized in that, include: A fixed stand is used to install the spin module, drying module, lateral vision inspection module, and axial vision inspection module, and is set at the fixed tool change point of the lower tool magazine; The spin module is used to fix the tool holder and rotate the tool during inspection to facilitate the acquisition of side and bottom edge images at different angles. The lateral vision inspection module is used to acquire images of the side edge of the tool and can move in the horizontal direction perpendicular to the rotation axis of the spin module, and is equipped with a first detection unit. The axial vision inspection module is used to acquire images of the bottom edge of the tool and can move in a direction parallel to the rotation axis of the spin module. The air-drying module, located between the spin module and the axial vision inspection module, is used to spray airflow toward the tool direction to remove cutting fluid and debris adhering to the tool to be inspected. It can move together with the axial vision inspection module and has a second inspection unit inside to control the movement position of the axial vision inspection module and the movement of the lateral vision inspection module. The control system controls the spin module, drying module, lateral vision inspection module, and axial vision inspection module. During inspection, the spin module rotates, the axial vision inspection module moves toward the tool, and the drying module is activated. When the second inspection unit detects the tool, the axial vision inspection module stops moving, ensuring that the bottom edge of the tool is within the imaging depth of the axial vision inspection module. Simultaneously, the lateral vision inspection module begins moving. When the first inspection unit detects the tool, the lateral vision inspection module stops moving, ensuring that the side edge of the tool is within the imaging depth of the lateral vision inspection module.

7. The tool magazine tool detection device according to claim 6, characterized in that, The spin module includes a servo motor, a reducer, a rear cover of the motor mount, a tool holder connector, a front cover of the motor mount, a rotation module support plate, spring plungers, and a tool holder. The spin module is driven by the servo motor connected to the reducer. The output end of the servo motor is connected to the input end of the reducer and fixed to the rear cover of the motor mount. One end of the tool holder connector is connected to the output end of the reducer, and the other end is fixed to the tool holder. The tool holder connector is rotatably connected to the front cover of the motor mount. The rear cover of the motor mount is connected to the front cover of the motor mount. The front cover of the motor mount is fixed to the rotation module support plate, and the rotation module support plate is fixed to a fixed frame. Four spring plungers are evenly distributed at the end of the tool holder for clamping the tool holder pull studs.

8. The tool magazine tool detection device according to claim 6, characterized in that, The air-drying module includes a through-beam sensor for axial tool position detection, a high-speed airflow cavity, an air vent, and a fixed bracket fixture. The high-speed airflow cavity is an annular structure with an internal airflow cavity. Multiple air slits are opened on the inner circular surface along the feed direction of the tool to be tested. The through-beam sensor for axial tool position detection is symmetrically installed in the high-speed airflow cavity. The side end is fixed to the bracket fixture by an adjustable sliding clamp. The fixed bracket fixture is adjustable in the vertical direction.

9. The tool magazine tool detection device according to claim 6, characterized in that, The lateral vision detection module includes a first moving unit, a first slide block, a diffuse reflection sensor for the lateral vision module, a first fine-tuning slide, a side blade detection light source, a side blade detection lens, and a side blade detection camera; the first slide block moves with the first moving unit; the diffuse reflection sensor is fixed to the front end of the first slide block, the first fine-tuning slide block is fixed to the rear end of the first slide block, and the side blade detection light source, the side blade detection lens, and the side blade detection camera are arranged from front to back on the first fine-tuning slide block.

10. The tool magazine tool detection device according to claim 6, characterized in that, The axial vision inspection module includes a second moving unit, a second slide block, a second fine-tuning slide table, a bottom edge detection light source, a bottom edge detection lens, a bottom edge detection camera, and a support plate; the second moving unit is fixed on the support plate, the second slide block moves with the second moving unit, the second fine-tuning slide table is fixed to the rear end of the second slide block, and the side edge detection light source, the side edge detection lens, and the side edge detection camera are arranged from front to back on the second fine-tuning slide table; the support plate is fixed to the fixed frame.

Images