Device and method for detecting size of part of small power transmission device

Abstract

The invention relates to a small power transmission device part size detection device and method. The device is composed of a motion control system, an image acquisition system, a detection tool, an image processing system and a computer control system. The motion control system comprises a two-axis motion sliding table and a fixed bottom plate. The image acquisition system comprises an industrial camera, a telecentric lens and a coaxial refraction light source. The detection tool is composed of a mechanical structural part suitable for the cylindrical part and a material disc. The image processing system includes developing size detection algorithms for different parts. The computer control system comprises a computer and control software and a motion controller in the computer. The device can meet the size detection requirements of different parts, and has the characteristics of non-contact, high efficiency, high precision and automation.

Description

Technical Field

[0001] The invention relates to a device and method for detecting the size of parts of a small power transmission device, and belongs to the field of machine vision and precision detection. Background Art

[0002] In ballistic missiles and launch vehicles, inertial navigation systems play a very important role, ensuring that missiles (or rockets) hit targets (or enter orbit accurately) as per predetermined requirements. Inertial devices and inertial instruments are key parts of inertial navigation systems, and their performance directly affects the accuracy of missiles (or rockets) hitting targets (or entering orbit accurately). Inertial devices and inertial instruments are high-precision components, which are composed of several precision parts. Small power transmission devices are important components in inertial instruments, and their parts have high dimensional matching accuracy requirements, which require dimensional inspection and selection before assembly.

[0003] At present, the dimensional inspection of small power transmission devices still relies on experienced inspectors, who use plug gauges and high-precision rulers under a microscope to observe the dimensions of parts before assembly, and select parts according to the dimensions. This process is highly dependent on personnel experience, and the repeatability, reliability and accuracy are all poor, which to some extent affects the assembly quality and production efficiency of small power transmission devices, and affects the accuracy of the inertial navigation system. Summary of the invention

[0004] The technical problem solved by the present invention is: to overcome the shortcomings of the prior art and provide a device and method for detecting the size of parts of a small power transmission device, aiming to complete the size detection of four types of key parts, namely cylindrical parts, filamentary parts, ceramic parts and brush parts, through automatic detection technology, and to improve the efficiency and quality of parts selection through non-contact, high-efficiency and high-precision automatic size detection, thereby fundamentally improving the quality of parts after assembly and improving the accuracy and stability of the inertial navigation system.

[0005] The technical solution of the present invention is: a device for detecting the size of parts of a small power transmission device, the device comprising a motion control system, an image acquisition system, a detection tool, an image processing system and a computer control system;

[0006] The motion control system, driven by the motion control signal, moves the image acquisition system and the inspection tooling so that the inspected parts fall into the center of the field of view of the image acquisition system and keep the image clear;

[0007] Image acquisition system, used to obtain images of inspected parts and control image quality;

[0008] Inspection tooling, used to fix the inspected parts;

[0009] The image processing system is used to process the collected part images, run the size detection algorithm, obtain the size detection results and send them to the computer control system;

[0010] The computer control system is used for controlling the image acquisition system, operating the motion control system, displaying and saving images in real time, and displaying and saving size detection results.

[0011] Preferably, the motion control system comprises: a vertical slide, a horizontal slide and a slide mounting base plate;

[0012] The horizontal slide is connected to the slide installation base plate, and the horizontal slide can move along the horizontal direction relative to the slide installation base plate;

[0013] The vertical slide is connected to the slide mounting base plate, and the vertical slide can move along the vertical direction relative to the slide mounting base plate;

[0014] There is no mechanical connection between the horizontal slide and the vertical slide.

[0015] Preferably, the image acquisition system comprises an industrial camera, a telecentric lens, a support frame, an annular buckle, an upper coaxial refractive light source and a lower coaxial refractive light source;

[0016] The industrial camera is connected to the telecentric lens, the annular buckle is installed on the upper side of the telecentric lens and clamps the lens, the annular buckle is connected to the support frame through screws, and the support frame is fixed to the vertical slide through screws;

[0017] Industrial cameras and telecentric lenses can move with the vertical slide;

[0018] The upper coaxial refraction light source is fixed to the mechanical bottom plate of the vertical slide by screws and does not move with the vertical slide;

[0019] The lower coaxial refraction light source is fixed on the slide mounting bottom plate by screws.

[0020] Preferably, the detection tooling includes a mechanical structure for fixing cylindrical parts, and a material tray for fixing filamentary parts, ceramic parts and brush parts; the detection tooling is fixed on a horizontal slide.

[0021] Preferably, the material tray is mounted on the top of the horizontal slide using screws, with a through hole in the middle for placing the filamentary parts flatly above the through hole, multiple parallel bosses on both sides, and a groove between the two bosses for placing the base of the ceramic parts or brush parts so that the ceramic parts or brush parts are placed flat on the bosses.

[0022] Preferably, the upper coaxial refraction light source and the lower coaxial refraction light source are started separately according to the different parts to be detected. When detecting the size of cylindrical parts, the lower coaxial refraction light source is used, and when detecting filamentary parts, ceramic parts and brush parts, the upper coaxial refraction light source is used. The upper coaxial refraction light source and the lower coaxial refraction light source are not started at the same time.

[0023] Preferably, the optical paths of the upper coaxial refractive light source and the lower coaxial refractive light source are refractive optical paths, and the light source is a reflector in the vertical direction, so that light can be transmitted normally; the image acquisition system located above the light source can capture the image of the part through the upper coaxial refractive light source.

[0024] Preferably, when detecting the size of a cylindrical part, the lower coaxial refractive light source illuminates from the back of the cylindrical part, and the image acquisition system acquires the image as the shadow of the part; when detecting filamentary parts, ceramic parts and brush parts, the upper coaxial refractive light source illuminates from above the part, and the image acquisition system acquires the image as the surface reflection image of the part.

[0025] Preferably, the size detection algorithm comprises the following steps:

[0026] S1. Use Gaussian filtering algorithm to equalize the image background noise, and then use adaptive threshold processing algorithm to reduce the background noise and retain the contour feature information of the detected parts in the image;

[0027] S2, using the contour tracking algorithm, using the edge contour feature information of the detected part, calculating the number of part contour points, and obtaining the position set of the part size contour points in the image, which is recorded as the contour tracking original point set;

[0028] S3. Calculate the geometrical dimension characteristics of the part by using the geometrical feature fitting method according to the number of contour points and the two-dimensional positions of the contour points;

[0029] S4, comparing the fitting points on the fitted geometric features with the original points obtained by contour tracking, and removing noise points from the original points of contour tracking; when the position difference between a fitting point and the corresponding original point is greater than a set threshold, the original point is determined to be a noise point;

[0030] S5. Use the contour with some noise removed to track the original point set, and iteratively execute steps S3 to S5 N times to remove more noise and obtain a more accurate size detection result, where N is greater than or equal to 5.

[0031] Another technical solution of the present invention is: a part size detection method, the method comprising the following steps:

[0032] Depending on the type of part to be inspected, the operator fixes the part on the mechanical structure or tray;

[0033] Adjust the position of the image acquisition system and the tooling until a clear part image is obtained and the part is located in the center of the image;

[0034] A size detection algorithm is used to perform size detection on the collected part image to obtain the size of the part; the size detection algorithm includes the following steps:

[0035] S1. Use Gaussian filtering algorithm to equalize the image background noise, and then use adaptive threshold processing algorithm to reduce the background noise and retain the contour feature information of the detected parts in the image;

[0036] S2, using the contour tracking algorithm, using the edge contour feature information of the detected part, calculating the number of part contour points, and obtaining the position set of the part size contour points in the image, which is recorded as the contour tracking original point set;

[0037] S3. Calculate the geometrical dimension characteristics of the part by using the geometrical feature fitting method according to the number of contour points and the two-dimensional positions of the contour points;

[0038] S4, comparing the fitting points on the fitted geometric features with the original points obtained by contour tracking, and removing noise points from the original points of contour tracking; when the position difference between a fitting point and the corresponding original point is greater than a set threshold, the original point is determined to be a noise point;

[0039] S5. Use the contour with some noise removed to track the original point set, and iteratively execute steps S3 to S5 N times to remove more noise and obtain a more accurate size detection result, where N is greater than or equal to 5.

[0040] The beneficial effects of the present invention compared with the prior art are:

[0041] (1) The present invention uses a computer to control the motion system, image acquisition system and image processing system to complete machine vision detection of part dimensions according to the characteristics of the parts, and in this process achieves a detection accuracy of 0.005 mm;

[0042] (2) The detection device of the present invention can effectively improve the detection efficiency and accuracy of part dimensions and improve the quality of assembled products. It has the advantages of non-contact, high precision, high efficiency and automation. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 This is a schematic diagram of a device for detecting the size of parts of a small power transmission device according to an embodiment of the present invention:

[0044] Figure 1Middle: 1-communication cable; 2-lower coaxial refraction light source; 3-mechanical structure; 4-material tray; 5-horizontal slide; 6-vertical slide; 7-buckle; 8-telecentric lens; 9-upper coaxial refraction light source; 10-support frame; 11-industrial camera; 12-slide mounting base; -13 computer.

[0045] Figure 2 It is a schematic diagram of the components of a device for detecting the size of parts of a small power transmission device according to an embodiment of the present invention;

[0046] Figure 3 This is a schematic diagram of a mechanical structure for fixing a cylindrical part according to an embodiment of the present invention:

[0047] Figure 4 This is a schematic diagram of a material tray according to an embodiment of the present invention:

[0048] Figure 5 is a control flow chart of the device according to an embodiment of the present invention;

[0049] Figure 6 It is a flow chart of the algorithm used by the image processing system of the embodiment of the present invention. DETAILED DESCRIPTION

[0050] The present invention is further described below in conjunction with examples and drawings, but the protection scope of the present invention should not be limited thereto.

[0051] like Figure 1 and Figure 2 As shown, the small power transmission device part size detection device provided by the present invention consists of five parts: a motion control system, an image acquisition system, a special detection tooling, an image processing system and a computer control system.

[0052] The motion control system, driven by the motion control signal, moves the image acquisition system and the inspection tooling so that the inspected parts fall into the center of the field of view of the image acquisition system and keep the image clear;

[0053] Image acquisition system, used to obtain images of inspected parts and control image quality;

[0054] Inspection tooling, used to fix the inspected parts;

[0055] The image processing system is used to process the collected part images, run the size detection algorithm, obtain the size detection results and send them to the computer control system;

[0056] The computer control system is used for controlling the image acquisition system, operating the motion control system, displaying and saving images in real time, and displaying and saving size detection results.

[0057] The motion control system includes: a vertical slide 6, a horizontal slide 5 and a slide mounting base 12.

[0058] The motion control system comprises: a vertical slide 6, a horizontal slide 5 and a slide mounting base 12;

[0059] The horizontal slide 5 is connected to the slide mounting base plate 12, and the horizontal slide 5 can move in the horizontal direction relative to the slide mounting base plate 12;

[0060] The vertical slide 6 is connected to the slide mounting base plate 12, and the vertical slide 6 can move along the vertical direction relative to the slide mounting base plate 12;

[0061] There is no mechanical connection between the horizontal slide 5 and the vertical slide 6.

[0062] The image acquisition system includes an industrial camera 11, a telecentric lens 8, a support frame 10, an annular buckle 7, an upper coaxial refractive light source 9 and a lower coaxial refractive light source 2;

[0063] The industrial camera 11 is connected to the telecentric lens 8 using the standard C-mount interface provided by the product. The annular buckle 7 is installed on the upper side of the telecentric lens 8 and clamps the lens. The annular buckle 7 is connected to the support frame 10 by screws, and the support frame 10 is fixed to the vertical slide 6 by screws.

[0064] The industrial camera 11 and the telecentric lens 8 can move with the vertical slide 6;

[0065] The upper coaxial refractive light source 9 is fixed to the mechanical bottom plate of the vertical slide 6 by screws and does not move with the vertical slide 6;

[0066] The lower coaxial refractive light source 2 is fixed to the slide mounting base plate 12 by screws.

[0067] The detection tooling includes a mechanical structure 3 for fixing cylindrical parts, and a material tray 4 for fixing filamentary parts, ceramic parts and brush parts; the detection tooling is fixed on a horizontal slide 5.

[0068] like Figure 3 As shown, the mechanical structure 3 is a copper tire-holding structure, which is installed on the side wall of the horizontal slide 5 using screws, and the cylindrical parts are fixed in the form of a spiral tire.

[0069] like Figure 4As shown, the tray 4 includes installation positions for three types of parts. By replacing the clamping components of the tray 4, the filamentary parts, ceramic parts and brush parts can be fixed respectively. The tray 4 is installed on the top of the horizontal slide 5 with screws, and a through hole is provided in the middle. The top of the through hole is used to lay the filamentary parts flat. Multiple bosses parallel to each other are provided on both sides, and a groove is provided between the two bosses. The groove is used to place the base of the ceramic part or the brush part, so that the ceramic part or the brush part is placed flat on the boss. According to the different placement positions of the parts, the horizontal slide 5 can be used to adjust the detected parts to the center of the image to improve the detection accuracy.

[0070] The material tray 4 is made of black aluminum to reduce the background reflection of the image.

[0071] Through the above two special tools, four types of parts can be fixed.

[0072] The mechanical structure 3 of the cylindrical part is detachable. The mechanical structure 3 is designed with 4 fixed threaded holes. The horizontal slide 5 is correspondingly designed with 4 threaded holes. When the size of the cylindrical part needs to be detected, the operator can use screws to install the mechanical structure 3 and fix the cylindrical part through the spiral tire structure. After the detection is completed, the operator can remove the screws to disassemble the mechanical structure 3.

[0073] The upper coaxial refraction light source 9 and the lower coaxial refraction light source 2 are started separately according to the different parts to be detected. When detecting the size of cylindrical parts, the lower coaxial refraction light source 2 is used, and when detecting filamentary parts, ceramic parts and brush parts, the upper coaxial refraction light source 9 is used. The upper coaxial refraction light source 9 and the lower coaxial refraction light source 2 are not started at the same time.

[0074] The optical paths of the upper coaxial refractive light source 9 and the lower coaxial refractive light source 2 are refractive optical paths. The light source is a 45° reflector in the vertical direction, so it can transmit light normally. The image acquisition system located above the light source can capture the part image through the upper coaxial refractive light source 9.

[0075] When detecting the size of cylindrical parts, the lower coaxial refractive light source 2 illuminates from the back of the cylindrical parts, and the image acquisition system will acquire the image as the shadow of the part; when detecting filamentary parts, ceramic parts and brush parts, the upper coaxial refractive light source 9 illuminates from above the parts, and the image acquisition system will acquire the image as the surface reflection image of the part.

[0076] In the image acquisition system, the upper coaxial refractive light source 9 and the lower coaxial refractive light source 2 are installed on the mechanical bottom plate of the vertical slide 6, and do not move with the movement of the slide, and the position of the light source is fixed. Before acquiring the image, the distance between the industrial camera 11 and the telecentric lens 8 parts is adjusted, and after the optimal position is determined, the relative distance remains unchanged. The placement position of the telecentric lens (8) can be made to minimize the edge distortion of the image. The industrial camera (11) used in the device has a resolution of more than 20 million pixels and is used to acquire images of the parts on the material tray (4) or the mechanical structure 3. The vertical slide 6 and the horizontal slide 5 can move up and down and left and right, so that the industrial camera 11 and the parts reach the accurate position.

[0077] The computer control system includes a computer 13 , a communication cable 1 and device control software and a motion mechanism controller contained therein, and the image processing system runs in the computer 13 .

[0078] The computer control system includes a computer 13 and device control software and a motion mechanism controller contained therein.

[0079] The control software is used by operators to control the operation of the small power transmission device part size detection device using a computer, including device initialization, operation of the motion control system, operation of the image acquisition system, operation of the image processing system and result storage functions. Operators can also use the "one-click" control function in the control software to automatically complete related functions preset by the computer control system.

[0080] The motion mechanism controller is an accessory of the computer control system, used for decoding and outputting computer control signals. It is connected to the computer through a USB interface and to the vertical slide 6 and the horizontal slide 5 through an RS232 serial port, thereby converting the computer control signal into the final motion of the motion slide.

[0081] The software compiled in the computer control system includes functions such as image acquisition control, size detection algorithm, motion slide control, light source control, and real-time image display. The above functions are integrated into the same control software, and the operator does not need to control each hardware separately.

[0082] The computer control system and the image acquisition system are connected by a network cable, and the image information is transmitted using the TCP\IP protocol.

[0083] The computer control system includes an emergency stop function, which can implement an emergency stop on the vertical slide 6 and the horizontal slide 5 in the event of an unexpected situation. The emergency stop includes software emergency stop and hardware emergency stop. The control software in the computer control system will detect the real-time position of the motion control system. When the motion slide is at risk of position collision, the software emergency stop function will obtain the highest control priority, suspend all other running functions, and stop the motion control system. The hardware emergency stop is implemented by the motion mechanism controller included in the computer control system. When the operator finds that an unexpected situation has occurred in the device, he can directly press the emergency stop button on the motion controller to cut off the power supply of the motion slide to ensure the safety of the device.

[0084] like Figure 6 As shown, in the image processing system, the size detection algorithm includes the following steps:

[0085] S1. Use GaussianBlur algorithm to equalize the image background noise, and then use adaptive threshold processing algorithm to reduce the background noise and retain the contour feature information of the detected parts in the image;

[0086] S2, using the contour tracking algorithm (findContours algorithm), using the edge contour feature information of the detected part, calculating the number of part contour points, and obtaining the position set of the part size contour points in the image, which is recorded as the contour tracking original point set;

[0087] S3. According to the number of contour points and the two-dimensional position of contour points, the geometric feature fitting method is used to calculate the geometric dimension features of the parts; among them, the filamentary parts, ceramic parts and brush parts are fitted by the straight line fitting algorithm (fitline algorithm) and the minimum enclosing rectangle algorithm (minAreaRect algorithm), and the cylindrical parts are fitted by the ellipse fitting algorithm (fitEllipse algorithm). The above algorithms can obtain the geometric dimension features of the parts such as the side length, perimeter, center of the circle, and spatial inclination angle;

[0088] S4, comparing the fitting points on the fitted geometric features with the original points obtained by contour tracking, and removing noise points from the original points of contour tracking; when the position difference between a fitting point and the corresponding original point is greater than a set threshold, the original point is determined to be a noise point;

[0089] S5. Use the contour with some noise removed to track the original point set, and iteratively execute steps S3 to S5 N times to remove more noise and obtain a more accurate size detection result, where N is greater than or equal to 5.

[0090] In the above device, the computer control system sends a control signal to the motion control system to make the motion control system complete the required position movement, and then sends a control signal to the image acquisition system through the computer control system to turn on and off the corresponding light source and start the industrial camera to achieve image acquisition and storage. After completing the image acquisition, the computer control system displays the image in real time on the software, sends a signal to the image processing system, runs the size detection algorithm, obtains the size detection result value, and saves the detection result to the computer hard disk.

[0091] like Figure 5 As shown, based on the above device, the present invention also provides a part size detection method, which includes the following steps:

[0092] Depending on the type of part to be inspected, the operator fixes the part on the mechanical structure (3) or the tray 4;

[0093] Adjust the position of the image acquisition system and the tooling until a clear part image is obtained and the part is located in the center of the image;

[0094] Use the size detection algorithm to detect the size of the collected part image to obtain the size of the part;

[0095] Before the device works, the image acquisition system is calibrated first. The calibration method is Zhang's calibration method. The specific calibration process is as follows:

[0096] 1) Place a chessboard with a length of 9 grids and a width of 9 grids, a single grid size of 1.5*1.5mm, and a dimensional accuracy of 1μm at a position where it can be clearly photographed, and adjust the brightness to make the image gradient appropriate;

[0097] 2) Using an industrial camera 11 to take 12 chessboard photos with different placement angles and positions, and transferring them to a computer 13 for storage;

[0098] 3) Use Zhang's calibration algorithm written in computer 13 to process the image, and use sub-pixel corner detection, threshold processing, morphological processing and other methods to obtain the distortion coefficient and translation vector, so as to calibrate the internal parameters and distortion coefficient of the camera, and generate a calibration file. During the operation of the device, the calibration file is directly used without additional calibration.

[0099] 4) Recalibrate the device annually according to the above method.

[0100] After the calibration is completed, the device enters the working state.

[0101] Operators mainly perform the following operations to perform the test:

[0102] Determine the parts that need to be inspected, and distinguish between filamentary parts, ceramic parts, brush parts and cylindrical parts;

[0103] According to the selected part type, the tooling to be used is determined, and the filamentary parts, ceramic parts and brush parts are installed on the material tray 4, and the cylindrical parts are installed on the mechanical structure 3. When testing the filamentary parts, ceramic parts and brush parts, the mechanical structure 3 needs to be removed.

[0104] If the part to be measured is a cylindrical part, it is fixed on the mechanical structure 3, the lower coaxial refraction light source 2 is started, the upper coaxial refraction light source 9 is turned off, the vertical slide 6 is moved to adjust the position of the image acquisition system until a clear image of the part is obtained, and then the horizontal slide 5 is adjusted until the part is located at the center of the image, and then the image of the part is collected for size detection. If the part to be measured is a filamentary part, a ceramic part, or a brush part, the mechanical structure 3 is removed, the corresponding clamping assembly is selected on the tray 4 to fix the part, the lower coaxial refraction light source 2 is turned off, the upper coaxial refraction light source 9 is started, the vertical slide 6 is moved to adjust the position of the image acquisition system until a clear image of the part is obtained, and then the horizontal slide 5 is adjusted until the part is located at the center of the image, and then the image of the part is collected for size detection.

[0105] The operator starts the software, selects the specific type of the part to be inspected, enters the size information, and clicks the start inspection button. The control software will automatically call the preset program and adjust the vertical slide 6 and the horizontal slide 5 to enable the part to be imaged clearly. When the preset program cannot be imaged clearly due to part size deviation or installation deviation, the operator can fine-tune the position of the vertical slide 6 and the horizontal slide 5 through the manual setting function in the software to achieve clear imaging.

[0106] The software completes the detection through one-click operation. After the operator clicks, the image acquisition system and image processing system will automatically complete functions such as image acquisition, image preprocessing, threshold processing, contour tracking, noise elimination, geometric feature fitting, and size information calculation, and output the detection results to the computer hard disk for storage.

[0107] For the size of parts, this device mainly uses pre-processing methods such as Gaussian filtering and weighted median filtering. After pre-processing, image quality discrimination algorithms such as morphological processing, gradient analysis, and grayscale analysis are performed. After obtaining the optimal image, contour information is obtained by contour tracking and contour singular point removal, and geometric features such as center of circle, edge line, and cusp are fitted according to the contour information, and size calculation is completed.

[0108] In summary, the device can automatically complete the full-automatic, non-contact, high-efficiency and high-accuracy detection of two major categories and four types of parts, improve the efficiency of parts selection and assembly quality, and improve the final working effect of the inertial navigation system.

[0109] Although the present invention has been disclosed as above in the form of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications to the technical solution of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the protection scope of the technical solution of the present invention.

Claims

1. A device for detecting the size of parts of a small power transmission device, characterized in that Including motion control system, image acquisition system, detection tooling, image processing system and computer control system; The motion control system, driven by the motion control signal, moves the image acquisition system and the inspection tooling so that the inspected parts fall into the center of the field of view of the image acquisition system and keep the image clear; Image acquisition system, used to obtain images of inspected parts and control image quality; Inspection tooling, used to fix the inspected parts; The image processing system is used to process the collected part images, run the size detection algorithm, obtain the size detection results and send them to the computer control system; The computer control system is used for controlling the image acquisition system, operating the motion control system, displaying and saving images in real time, and displaying and saving size detection results.

2. A device for detecting the size of parts of a small power transmission device according to claim 1, characterized in that: The motion control system comprises: a vertical slide table (6), a horizontal slide table (5) and a slide table mounting base plate (12); The horizontal slide (5) is connected to the slide mounting base plate (12), and the horizontal slide (5) can move along the horizontal direction relative to the slide mounting base plate (12); The vertical slide (6) is connected to the slide mounting base plate (12), and the vertical slide (6) can move along the vertical direction relative to the slide mounting base plate (12); There is no mechanical connection between the horizontal slide table (5) and the vertical slide table (6).

3. A device for detecting the size of parts of a small power transmission device according to claim 2, characterized in that: The image acquisition system comprises an industrial camera (11), a telecentric lens (8), a support frame (10), an annular buckle (7), an upper coaxial refractive light source (9) and a lower coaxial refractive light source (2); The industrial camera (11) is connected to the telecentric lens (8), the annular buckle (7) is installed on the upper side of the telecentric lens (8) and clamps the lens, the annular buckle (7) is connected to the support frame (10) by screws, and the support frame (10) is fixed on the vertical slide (6) by screws; The industrial camera (11) and the telecentric lens (8) can move along with the vertical slide (6); The upper coaxial refractive light source (9) is fixed to the mechanical bottom plate of the vertical slide (6) by screws and does not move with the vertical slide (6); The lower coaxial refractive light source (2) is fixed on the slide mounting base plate (12) by means of screws.

4. A device for detecting the size of parts of a small power transmission device according to claim 2, characterized in that: The detection tooling comprises a mechanical structure (3) for fixing cylindrical parts, and a material tray (4) for fixing filamentary parts, ceramic parts and brush parts; the detection tooling is fixed on a horizontal slide table (5).

5. A device for detecting the size of parts of a small power transmission device according to claim 4, characterized in that: The material tray (4) is mounted on the top of a horizontal slide (5) using screws, with a through hole in the middle for placing filamentary parts flatly above the through hole, and multiple bosses parallel to each other are provided on both sides, with a groove between the two bosses for placing the base of a ceramic part or a brush part, so that the ceramic part or the brush part is placed flat on the boss.

6. A device for detecting the size of parts of a small power transmission device according to claim 4, characterized in that: The upper coaxial refraction light source (9) and the lower coaxial refraction light source (2) are respectively started according to the different parts to be detected. When detecting the size of cylindrical parts, the lower coaxial refraction light source (2) is used, and when detecting filamentary parts, ceramic parts and brush parts, the upper coaxial refraction light source (9) is used. The upper coaxial refraction light source (9) and the lower coaxial refraction light source (2) are not started at the same time.

7. A device for detecting the size of parts of a small power transmission device according to claim 6, characterized in that: The optical paths of the upper coaxial refractive light source (9) and the lower coaxial refractive light source (2) are refractive optical paths. The light source is a 45° reflector in the vertical direction, so that light can be transmitted normally. The image acquisition system located above the light source can capture the image of the part through the upper coaxial refractive light source (9).

8. The device for detecting the size of parts of a small power transmission device according to claim 6, characterized in that: When detecting the size of a cylindrical part, the lower coaxial refractive light source (2) illuminates the back of the cylindrical part, and the image acquisition system acquires the shadow of the part; when detecting a filamentary part, a ceramic part, or a brush part, the upper coaxial refractive light source (9) illuminates the part from above, and the image acquisition system acquires the surface reflection image of the part.

9. A device for detecting the size of parts of a small power transmission device according to claim 1, characterized in that: The size detection algorithm comprises the following steps: S1. Use Gaussian filtering algorithm to equalize the image background noise, and then use adaptive threshold processing algorithm to reduce the background noise and retain the contour feature information of the detected parts in the image; S2, using the contour tracking algorithm, using the edge contour feature information of the detected part, calculating the number of part contour points, and obtaining the position set of the part size contour points in the image, which is recorded as the contour tracking original point set; S3. Calculate the geometrical dimension characteristics of the part by using the geometrical feature fitting method according to the number of contour points and the two-dimensional positions of the contour points; S4, comparing the fitting points on the fitted geometric features with the original points obtained by contour tracking, and removing noise points from the original points of contour tracking; when the position difference between a fitting point and the corresponding original point is greater than a set threshold, the original point is determined to be a noise point; S5. Use the contour with some noise removed to track the original point set, and iteratively execute steps S3 to S5 N times to remove more noise and obtain a more accurate size detection result, where N is greater than or equal to 5.

10. A part size detection method according to claim 1, characterized in that The steps include: Depending on the type of parts to be inspected, the operator fixes the parts on the mechanical structure (3) or the material tray (4); Adjust the position of the image acquisition system and the tooling until a clear part image is obtained and the part is located in the center of the image; A size detection algorithm is used to perform size detection on the collected part image to obtain the size of the part; the size detection algorithm includes the following steps: S1. Use Gaussian filtering algorithm to equalize the image background noise, and then use adaptive threshold processing algorithm to reduce the background noise and retain the contour feature information of the detected parts in the image; S2, using the contour tracking algorithm, using the edge contour feature information of the detected part, calculating the number of part contour points, and obtaining the position set of the part size contour points in the image, which is recorded as the contour tracking original point set; S3. Calculate the geometrical dimension characteristics of the part by using the geometrical feature fitting method according to the number of contour points and the two-dimensional positions of the contour points; S4, comparing the fitting points on the fitted geometric features with the original points obtained by contour tracking, and removing noise points from the original points of contour tracking; when the position difference between a fitting point and the corresponding original point is greater than a set threshold, the original point is determined to be a noise point; S5. Use the contour with some noise removed to track the original point set, and iterate and execute steps S3 to S5 N times to remove more noise and obtain a more accurate size detection result, where N is greater than or equal to 5.

Images