A visual-based spindown measurement method, computer and measurement robot

Abstract

The application relates to a visual-based spindle speed measurement method, a computer and a measurement robot, and provides a visual-based spindle speed measurement method, which comprises the following steps: firstly, setting a photographing frequency of a photo; then, recognizing and detecting the content on the photo by using a target detection model; and finally, calculating the rotating speed of a spindle by using the position change amount of a wire passing hole. Compared with the existing method of detecting the rotating speed of the spindle by using a sensor, the method has lower equipment cost, and compared with the manual detection method of the rotating speed of the spindle, the method has higher measurement accuracy and is not affected by the errors caused by manual detection.

Description

Technical Field

[0001] This invention relates to the field of textile automation technology, and more specifically, to a vision-based spindle speed measurement method, a computer, and a measurement robot. Background Technology

[0002] A doubling twister is a twisting device that can achieve two twists in one rotation, significantly increasing twisting efficiency compared to traditional twisting equipment. It can bond two or more single yarns together to form a single strand, enhancing the properties of the original yarn. The doubling twist spindle is the heart and key component of the doubling twister; the biggest difference between a doubling twister and a ring twister lies in the spindle. A regular spindle adds only one twist per rotation, while a doubling twist spindle produces two twists per rotation.

[0003] Currently, the chemical fiber filament weaving industry commonly uses double-layer filament twisting machines. These machines have two layers and two sides, with each layer typically containing 128 spindles, each representing a production unit. Both the upper and lower layers have a long belt. The spindles rotate through friction between the belt and the spindle feet. This means that all 128 spindles in each layer are powered by the same belt. If the belt tension becomes inconsistent due to temperature, aging, or other factors, resulting in inconsistent pressure and friction between the belt and each spindle, it will cause spindle speed deviations. These deviations lead to uneven twist in the product, ultimately causing fabric defects during the weaving process. Currently, the standard speed for domestic double-twisting machines is set at approximately 11,000 rpm, or 183.3 spindle revolutions per second. Due to the unavoidable slippage and uneven friction caused by belt drives, the double-twisting machine, as a critical process, can cause quality problems such as strong twist, weak twist, and uneven twist. Therefore, spindle speed monitoring is a crucial quality management action in twisting weaving factories.

[0004] In existing technology, one method for measuring the speed of spindles in a doubling twister is to install a speed sensor on each spindle. A Chinese utility model patent (application number: CN201520258151.7) proposes an online sensor that integrates spindle speed, twist, and yarn breakage detection. This device is a photoelectric ring sensor installed above the guide wheel of each spindle to detect the spindle speed or yarn speed. The disadvantages of this approach are the high cost of the sensors. For example, with 360 doubling twisters, each with 256 spindles, a total of 92,160 sensors would be required. At an average cost of 40 yuan per sensor, this would amount to 3.686 million yuan. Furthermore, these sensors are large, causing significant interference with daily operations after installation. Inserting or removing tubes can also impact the sensors, potentially damaging them. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a vision-based spindle speed measurement method, computer and measurement robot, so as to overcome the disadvantages of the existing technology that using sensors to measure spindle speed will increase equipment costs and affect equipment use.

[0006] The above-mentioned technical objective of the present invention is achieved through the following technical solution: a vision-based spindle speed measurement method, comprising:

[0007] Obtain the set rotation speed of the spindle of the doubling machine, calculate the first shooting frequency based on the set rotation speed, and take a first photo and a second photo of the same spindle according to the first shooting frequency;

[0008] The first image is input into a pre-trained target detection neural network model for detection, and the first position information corresponding to the spindle in the first image and the second position information corresponding to the wire passage hole in the first image are obtained.

[0009] The second photo is input into a pre-trained target detection neural network model for detection, and the third position information corresponding to the spindle in the second photo and the fourth position information corresponding to the wire passage hole in the second photo are obtained.

[0010] Based on the first and second position information, calculate the first angle between the line connecting the wire hole and the spindle axis in the first photograph and the preset coordinate axis; based on the third and fourth position information, calculate the second angle between the line connecting the wire hole and the spindle axis in the second photograph and the preset coordinate axis.

[0011] The first angle difference is calculated based on the first included angle and the second included angle; the first time difference between the first and second photos is calculated based on the first shooting frequency; and the first actual rotational speed of the spindle is calculated based on the first angle difference and the first time difference.

[0012] Optionally, obtaining the set rotational speed of the spindle of the doubling twister and calculating the first shooting frequency based on the set rotational speed includes:

[0013] Obtain the set speed of the spindle of the doubling twister ;

[0014] Obtain the pre-set first undersampling parameters ;

[0015] According to the set speed and the first undersampling parameter Calculate the first shooting frequency .

[0016] Optionally, the step of inputting the first photograph into a pre-trained target detection neural network model for detection to obtain the first position information corresponding to the spindle in the first photograph and the second position information corresponding to the wire passage hole in the first photograph includes:

[0017] The first photo is input into a pre-trained target detection neural network model for detection, and a first bounding box for defining the spindle in the first photo and a second bounding box for defining the thread hole in the first photo are obtained.

[0018] Get the first horizontal coordinates of the top-left vertex of the first bounding box. and the first lateral length of the first bounding box ; set the first horizontal coordinate and the first horizontal length Record this as the first position information;

[0019] Get the second horizontal coordinates corresponding to the top left vertex of the second bounding box. and the second lateral length of the second bounding box ; set the second horizontal coordinate and the second lateral length This is recorded as the second position information.

[0020] Optionally, the step of calculating the first angle between the line connecting the wire guide hole and the spindle axis in the first photograph and the preset coordinate axis based on the first position information and the second position information includes:

[0021] Get the radius of the spindle ;

[0022] Obtain the fifth x-coordinate of the spindle's axis in the first image. :

[0023] ;

[0024] Obtain the sixth x-coordinate of the center point of the thread hole in the first image. :

[0025] ;

[0026] Calculate the first horizontal distance between the fifth and sixth horizontal coordinates. :

[0027] ;

[0028] According to the radius and the first lateral distance Calculate the first included angle :

[0029] .

[0030] Optionally, the step of inputting the second photograph into a pre-trained target detection neural network model for detection to obtain the third position information corresponding to the spindle in the second photograph and the fourth position information corresponding to the wire passage hole in the second photograph includes:

[0031] The second photo is input into a pre-trained target detection neural network model for detection, and a third bounding box for defining the spindle in the second photo and a fourth bounding box for defining the thread hole in the second photo are obtained accordingly.

[0032] Get the third horizontal coordinates of the top-left vertex of the third bounding box. And the third horizontal length of the third bounding box ; the third horizontal coordinate and the third horizontal length Recorded as third position information;

[0033] Get the fourth horizontal coordinate of the top-left vertex of the fourth bounding box. And the fourth horizontal length of the fourth bounding box ; the fourth horizontal coordinate and the fourth horizontal length This is recorded as the fourth position information.

[0034] Optionally, the step of calculating the second included angle between the line connecting the wire guide hole and the spindle axis in the second photograph and the preset coordinate axis based on the third and fourth position information includes:

[0035] Get the radius of the spindle ;

[0036] Obtain the seventh x-coordinate of the spindle's axis in the second image. :

[0037] ;

[0038] Obtain the eighth x-coordinate of the center point of the thread hole in the second photo. :

[0039] ;

[0040] Calculate the seventh x-coordinate With the eighth x-axis The second lateral distance between :

[0041] ;

[0042] According to the radius and the second lateral distance Calculate the second included angle :

[0043] .

[0044] Optionally, the step of calculating the first angle difference based on the first included angle and the second included angle, and calculating the first time difference between the first and second photos based on the first shooting frequency; and calculating the first actual rotational speed of the spindle based on the first angle difference and the first time difference, includes:

[0045] According to the first included angle Second angle Calculate the first angle difference ;

[0046] According to the first shooting frequency The first time difference between the first and second photos is calculated accordingly. ;

[0047] Based on the first angle difference and the first time difference Calculate the first actual rotational speed of the spindle. .

[0048] Optionally, it also includes: obtaining a pre-set second undersampling parameter. ;

[0049] According to the set speed and the second undersampling parameter Calculate the second shooting frequency :

[0050] ;

[0051] According to the second shooting frequency Take a third and fourth photograph of the same ingot;

[0052] The third photo is input into a pre-trained target detection neural network model for detection, and the fifth position information corresponding to the spindle in the third photo and the sixth position information corresponding to the wire passage hole in the third photo are obtained.

[0053] The fourth photo is input into a pre-trained target detection neural network model for detection, and the seventh position information corresponding to the spindle in the fourth photo and the eighth position information corresponding to the wire hole in the fourth photo are obtained.

[0054] Based on the fifth and sixth position information, calculate the third included angle between the line connecting the wire hole and the spindle axis in the third photograph and the preset coordinate axis. Based on the seventh and eighth position information, calculate the fourth included angle between the line connecting the wire guide hole and the spindle axis in the fourth photograph and the preset coordinate axis. ;

[0055] According to the third included angle and the fourth angle Calculate the second angle difference ;

[0056] According to the second shooting frequency The second time difference between the first and second photos is calculated accordingly. ;

[0057] Based on the second angle difference and the second time difference Calculate the second actual rotational speed of the spindle. ;

[0058] Determine the first actual speed Is it equal to the second actual speed? If so, then the first actual rotational speed will be... Or the second actual speed Record as actual speed .

[0059] A computer device includes a memory and a processor; the memory is connected to the processor, the memory is used to store a computer program, and the processor is used to invoke the computer program to cause the computer device to perform the method described above.

[0060] A vision-based spindle speed measurement robot includes: an AGV base, a column, a camera for photographing the spindle, a stroboscope for illumination, a battery, and the aforementioned computer equipment; the bottom of the column is fixedly connected to the AGV base; the battery is fixedly connected to the AGV base; the stroboscope and the camera are both fixedly connected to the side wall of the column; the battery is electrically connected to the camera, the stroboscope, the AGV base, and the computer equipment respectively.

[0061] In summary, the present invention has the following beneficial effects: This application provides a vision-based spindle speed measurement method. First, it sets the shooting frequency of photos, then uses a target detection model to identify and detect the content in the photos, and finally calculates the spindle rotation speed using the position change of the wire guide hole. Compared with existing methods that use sensors to detect spindle speed, this method has lower equipment costs and does not affect the use of the twisting machine. Furthermore, compared with manual detection of spindle speed, it has higher measurement accuracy and is not affected by errors caused by manual detection. Attached Figure Description

[0062] Figure 1 This is a flowchart of a vision-based ingot speed measurement method according to the present invention;

[0063] Figure 2 This is a structural diagram of a vision-based ingot speed measurement system according to the present invention;

[0064] Figure 3 This is an internal structural diagram of the computer device in an embodiment of the present invention;

[0065] Figure 4 A schematic diagram illustrating the measurement of spindle rotation speed using a stroboscope in existing technology;

[0066] Figure 5 This is a schematic diagram illustrating the use of a target detection neural network model to mark the spindle and the wire passage in this invention;

[0067] Figure 6 This is a schematic diagram illustrating the calculation of the first lateral distance according to the present invention;

[0068] Figure 7 This is a schematic diagram illustrating the calculation of the second lateral distance according to the present invention;

[0069] Figure 8 This is a spectrum diagram showing the position of the wire hole measured in this invention;

[0070] Figure 9 This is a spectrum diagram illustrating the difference between the actual rotational speed and the set rotational speed of the present invention.

[0071] Figure 10 This is a schematic diagram of the structure of a vision-based ingot speed measurement robot according to the present invention;

[0072] Figure 11 This is a schematic diagram of the vision-based spindle speed measurement robot of the present invention measuring the spindle rotation speed.

[0073] In the diagram: 1. Photo capture module; 2. Position information detection module; 3. Angle calculation module; 4. Spindle rotation speed calculation module. Detailed Implementation

[0074] To make the objectives, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Several embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein.

[0075] In this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances. The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature.

[0076] In this invention, unless otherwise expressly specified and limited, "above" or "below" a second feature can include direct contact between the first and second features, or contact between the first and second features through another feature between them. Furthermore, "above," "over," and "on top" of a second feature includes the first feature being directly above or diagonally above the second feature, or simply indicates that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" of a second feature includes the first feature being directly below or diagonally below the second feature, or simply indicates that the first feature is at a lower horizontal level than the second feature. The terms "vertical," "horizontal," "left," "right," "above," "below," and similar expressions are for illustrative purposes only and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed or operated in a specific orientation, and therefore should not be construed as limiting the invention.

[0077] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. Example 1

[0078] This invention provides a vision-based method for measuring spindle speed, such as... Figure 1 As shown, it includes:

[0079] S1. Obtain the set rotation speed of the spindle of the doubling machine, calculate the first shooting frequency based on the set rotation speed, and take a first photo and a second photo of the same spindle according to the first shooting frequency;

[0080] S2. Input the first photo into the pre-trained target detection neural network model for detection, and obtain the first position information corresponding to the spindle in the first photo and the second position information corresponding to the wire passage hole in the first photo.

[0081] S3. Input the second photo into the pre-trained target detection neural network model for detection, and obtain the third position information corresponding to the spindle in the second photo and the fourth position information corresponding to the wire passage hole in the second photo.

[0082] S4. Based on the first position information and the second position information, calculate the first angle between the line connecting the wire hole and the spindle axis in the first photograph and the preset coordinate axis; based on the third position information and the fourth position information, calculate the second angle between the line connecting the wire hole and the spindle axis in the second photograph and the preset coordinate axis.

[0083] S5. Calculate the first angle difference based on the first included angle and the second included angle; calculate the first time difference between the first and second photos based on the first shooting frequency; calculate the first actual rotational speed of the spindle based on the first angle difference and the first time difference.

[0084] Specifically, in actual production work, the applicant discovered that the aluminum cup of the spindle has a regular structure revolving around a pivot axis. When it rotates, it also rotates around the axis without translation. Therefore, the outline shape of the aluminum cup is the same whether it is rotating or stationary. Consequently, when the spindle rotates at high speed during production, it is difficult to discern the rotational speed or status of the spindle with the naked eye. To solve this problem, the applicant found that the aluminum cup of the spindle has thread-passing holes on its side wall. When the aluminum cup rotates one revolution, the thread-passing holes also rotate one revolution with the aluminum cup, maintaining the same rotational speed. Therefore, the applicant conceived of using a stroboscope to measure the speed of the aluminum cup, with the specific solution as follows:

[0085] First, obtain the preset rotational speed of the twisting machine spindle. For example, the current spindle speed is typically 11160 rpm (revolutions per minute), or 186 revolutions per second. Adjust the stroboscope to the same frequency as this rotational speed, i.e., 186 Hz. If the actual speed of the spindle matches the preset speed, that is, the frequency of the stroboscope, then after each flash of the stroboscope, the observed position of the wire guide hole relative to the aluminum cup should remain consistent, utilizing the persistence of vision. If there is a difference between the spindle's rotational speed and the preset speed, the wire guide hole will gradually shift at that flashing frequency. Generally, the speed of the wire guide hole's displacement is affected by the frequency difference. Since the spindle is driven by a belt, the spindle speed usually will not exceed the preset speed. Therefore, if the wire guide hole is observed to move within a predetermined time, it indicates that there is a difference between the spindle speed and the preset speed. However, this method relies too heavily on people's subjective judgment. Different people have different sensitivities to the position of the wire guide hole. Therefore, it is impossible to form a quantitative standard to judge whether the wire guide hole has shifted or whether the shift of the wire guide hole exceeds the standard, making it difficult to accurately judge whether the stator speed is less than the threshold.

[0086] To address the shortcomings of the aforementioned solutions, this application further proposes a vision-based method for measuring spindle speed, such as... Figure 5 , 6 As shown in Figure 7, existing methods utilize the wire guide hole on the spinning spindle as a reference point to represent the spindle's rotational speed. When the frequency of the stroboscope is adjusted to match the spindle's rotational speed, the wire guide hole appears to remain in one position due to visual persistence. Similarly, when photographing the spindle, if the camera frequency matches the spindle's rotational speed, the position of the wire guide hole in both photographs should be identical. Therefore, in actual speed measurement, the camera's shooting frequency needs to be consistent with the spindle's preset rotational frequency. Two photographs of the spindle are then taken. If the actual spindle speed differs from the preset speed, the position of the wire guide hole in the two photographs will differ. The change in the wire guide hole's angle is calculated based on this difference in position. Then, the total change in the wire guide hole's angle per unit time is calculated based on this change in angle and compared with the preset rotational speed. By performing an addition calculation, the actual rotational speed of the spindle can be obtained. .

[0087] For example, a typical twisted spindle rotates at 11,000 rpm. In 0.1 seconds, the spindle rotates 18 times. If a high-speed camera is used to photograph the spindle, the shooting time is plotted on the x-axis, and the percentage "K" representing the distance L from the wire-passing hole of the aluminum cup to the "zero point" relative to the diameter d is plotted on the y-axis. The resulting curve would look like this: Figure 8 As shown.

[0088] If there is a deviation between the spindle speed and the stroboscope speed, the rotation of the wire guide hole can be observed with the naked eye. The rotation speed is the difference between the actual speed and the speed set by the stroboscope. Assuming the standard spindle speed is 11160 rpm and the actual spindle speed is 10980 rpm, the deviation is 180 rpm. Setting the stroboscope frequency to 11160 rpm means the flash frequency is 11160 / 60 = 186 Hz. The effect observed with the naked eye is as follows... Figure 9 As shown ( Figure 9 The time length of the horizontal axis is 0.1 seconds, meaning the position of the wire guide hole will gradually shift. Therefore, by measuring the shift of the wire guide hole, calculating the speed difference between the wire guide hole and the standard speed, and finally adding the speed difference to the set speed, the actual rotational speed of the spindle can be obtained. Those skilled in the art will understand that the sign of the speed difference can be positive or negative; that is, the actual rotational speed of the spindle can be greater than or less than the set speed. It is only necessary to set the corresponding positive or negative sign and add it to the standard speed. This part will not be elaborated further in this application.

[0089] In summary, this application provides a vision-based spindle speed measurement method. First, the shooting frequency of the photos is set. Then, the content in the photos is identified and detected using a target detection model. Finally, the rotational speed of the spindle is calculated using the positional change of the wire-passing hole. Compared with the existing method of detecting spindle speed using sensors, this method has lower equipment costs. Moreover, compared with the method of manually detecting spindle speed, it has higher measurement accuracy and is not affected by errors caused by manual detection.

[0090] Furthermore, the step of obtaining the set rotational speed of the spindle of the doubling twister and calculating the first shooting frequency based on the set rotational speed includes: obtaining the set rotational speed of the spindle of the doubling twister. ; Obtain the pre-set first undersampling parameters According to the set rotation speed and the first undersampling parameter Calculate the first shooting frequency .

[0091] In practical applications, assuming a standard spindle speed of 11160 rpm (186 Hz) and an actual spindle speed of 10980 rpm (183 Hz), if a camera takes pictures at 186 fps (equivalent to the spindle's preset rotation speed), the sampling rate will be too high, requiring too many photos for AI recognition. This places high demands on both the camera and AI computing power, resulting in very high costs. According to the Nyquist sampling theorem, to recover the original signal without distortion from the sampled signal, the sampling frequency should be greater than twice the highest frequency of the signal. When the sampling frequency is less than twice the highest frequency of the spectrum, the signal spectrum exhibits aliasing. When the sampling frequency is greater than twice the highest frequency of the spectrum, the signal spectrum does not exhibit aliasing. Therefore, depending on the actual situation, the camera's shooting frequency can be reduced, i.e., the photos can be downsampled, reducing the camera's sampling frequency to one-fifth of the original frequency (the first undersampling parameter). Therefore, 186 / 5 = 37.2Hz, so the actual measured spindle speed is 183 / 5 = 36.6Hz, meaning there's a difference of 0.6Hz. After calculating the corresponding spindle speed difference, it's necessary to multiply the speed difference by the sampling rate to obtain the final spindle speed. .

[0092] Furthermore, the step of inputting the first photograph into a pre-trained target detection neural network model for detection, to obtain the first position information corresponding to the spindle in the first photograph and the second position information corresponding to the wire passage hole in the first photograph, includes:

[0093] The first photo is input into a pre-trained target detection neural network model for detection, and a first bounding box for defining the spindle in the first photo and a second bounding box for defining the thread hole in the first photo are obtained.

[0094] Get the first horizontal coordinates of the top-left vertex of the first bounding box. and the first lateral length of the first bounding box ; set the first horizontal coordinate and the first horizontal length Record this as the first position information;

[0095] Get the second horizontal coordinates corresponding to the top left vertex of the second bounding box. and the second lateral length of the second bounding box ; set the second horizontal coordinate and the second lateral length This is recorded as the second position information.

[0096] Furthermore, the step of calculating the first angle between the line connecting the wire guide hole and the spindle axis in the first photograph and the preset coordinate axis based on the first position information and the second position information includes:

[0097] Get the radius of the spindle ;

[0098] Obtain the fifth x-coordinate of the spindle's axis in the first image. :

[0099] ;

[0100] Obtain the sixth x-coordinate of the center point of the thread hole in the first image. :

[0101] ;

[0102] Calculate the first horizontal distance between the fifth and sixth horizontal coordinates. :

[0103] ;

[0104] According to the radius and the first lateral distance Calculate the first included angle :

[0105] .

[0106] Furthermore, the step of inputting the second photograph into a pre-trained target detection neural network model for detection, to obtain the third position information corresponding to the spindle in the second photograph and the fourth position information corresponding to the wire passage hole in the second photograph, includes:

[0107] The second photo is input into a pre-trained target detection neural network model for detection, and a third bounding box for defining the spindle in the second photo and a fourth bounding box for defining the thread hole in the second photo are obtained accordingly.

[0108] Get the third horizontal coordinates of the top-left vertex of the third bounding box. And the third horizontal length of the third bounding box ; the third horizontal coordinate and the third horizontal length Recorded as third position information;

[0109] Get the fourth horizontal coordinate of the top-left vertex of the fourth bounding box. And the fourth horizontal length of the fourth bounding box ; the fourth horizontal coordinate and the fourth horizontal length This is recorded as the fourth position information.

[0110] Furthermore, the step of calculating the second included angle between the line connecting the wire guide hole and the spindle axis in the second photograph and the preset coordinate axis based on the third and fourth position information includes:

[0111] Get the radius of the spindle ;

[0112] Obtain the seventh x-coordinate of the spindle's axis in the second image. :

[0113] ;

[0114] Obtain the eighth x-coordinate of the center point of the thread hole in the second photo. :

[0115] ;

[0116] Calculate the seventh x-coordinate With the eighth x-axis The second lateral distance between :

[0117] ;

[0118] According to the radius and the second lateral distance Calculate the second included angle :

[0119] .

[0120] In practical applications, such as Figure 4 , 5 As shown in Figures 6 and 7, the wire-passing holes on the spindle can be identified by a pre-trained target detection neural network model, and the position of the wire-passing holes can be calculated by using the coordinates and dimensions of the bounding box in the photograph.

[0121] Further, the step of calculating the first angle difference based on the first included angle and the second included angle, and calculating the first time difference between the first and second photos based on the first shooting frequency; and calculating the first actual rotational speed of the spindle based on the first angle difference and the first time difference, includes:

[0122] According to the first included angle Second angle Calculate the first angle difference ;

[0123] According to the first shooting frequency The first time difference between the first and second photos is calculated accordingly. ;

[0124] Based on the first angle difference and the first time difference Calculate the first actual rotational speed of the spindle. .

[0125] Furthermore, it also includes: obtaining a pre-set second undersampling parameter. ;

[0126] According to the set speed and the second undersampling parameter Calculate the second shooting frequency :

[0127] ;

[0128] According to the second shooting frequency Take a third and fourth photograph of the same ingot;

[0129] The third photo is input into a pre-trained target detection neural network model for detection, and the fifth position information corresponding to the spindle in the third photo and the sixth position information corresponding to the wire passage hole in the third photo are obtained.

[0130] The fourth photo is input into a pre-trained target detection neural network model for detection, and the seventh position information corresponding to the spindle in the fourth photo and the eighth position information corresponding to the wire hole in the fourth photo are obtained.

[0131] Based on the fifth and sixth position information, calculate the third included angle between the line connecting the wire hole and the spindle axis in the third photograph and the preset coordinate axis. Based on the seventh and eighth position information, calculate the fourth included angle between the line connecting the wire guide hole and the spindle axis in the fourth photograph and the preset coordinate axis. ;

[0132] According to the third included angle and the fourth angle Calculate the second angle difference ;

[0133] According to the second shooting frequency The second time difference between the first and second photos is calculated accordingly. ;

[0134] Based on the second angle difference and the second time difference Calculate the second actual rotational speed of the spindle. ;

[0135] Determine the first actual speed Is it equal to the second actual speed? If so, then the first actual rotational speed will be... Or the second actual speed Record as actual speed .

[0136] In practical applications, to avoid missing detections of frequency deviations that are integer multiples of the sampling frequency (i.e., when the frequency deviation is an integer multiple of the sampling frequency, it will be unobservable after sampling), the first actual rotational speed is calculated. Next, another sampling rate needs to be set, reducing the sampling rate from one-fifth to one-quarter (that is, the second undersampling parameter). ), measure the spindle speed again, and record it as the second actual speed. And with the first actual speed The two measurements are compared. If the results are the same, the speed measurement is accurate. If they are different, the measurement is inaccurate and needs to be repeated.

[0137] like Figure 2 As shown, the present invention also provides a vision-based spindle speed measurement system, comprising:

[0138] Photo capture module: used to acquire the set rotation speed of the spindle of the twisting machine, calculate the first shooting frequency based on the set rotation speed, and take a first photo and a second photo of the same spindle according to the first shooting frequency;

[0139] Position information detection module: used to input the first photo into a pre-trained target detection neural network model for detection, to obtain the first position information corresponding to the spindle in the first photo and the second position information corresponding to the wire passage hole in the first photo; input the second photo into the pre-trained target detection neural network model for detection, to obtain the third position information corresponding to the spindle in the second photo and the fourth position information corresponding to the wire passage hole in the second photo;

[0140] Angle calculation module: used to calculate the first angle between the line connecting the wire hole and the spindle axis in the first photograph and the preset coordinate axis based on the first position information and the second position information; and to calculate the second angle between the line connecting the wire hole and the spindle axis in the second photograph and the preset coordinate axis based on the third position information and the fourth position information.

[0141] Spindle rotation speed calculation module: Calculates the first angle difference based on the first included angle and the second included angle; calculates the first time difference between the first and second photos based on the first shooting frequency; calculates the first actual rotation speed of the spindle based on the first angle difference and the first time difference.

[0142] For specific limitations regarding a vision-based ingot speed measurement system, please refer to the limitations of a vision-based ingot speed measurement method described above, which will not be repeated here. Each module in the aforementioned vision-based ingot speed measurement system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0143] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 3 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computational and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage medium. When executed by the processor, the computer program implements a vision-based ingot speed measurement method.

[0144] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0145] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps: including:

[0146] Obtain the set rotation speed of the spindle of the doubling machine, calculate the first shooting frequency based on the set rotation speed, and take a first photo and a second photo of the same spindle according to the first shooting frequency;

[0147] The first image is input into a pre-trained target detection neural network model for detection, and the first position information corresponding to the spindle in the first image and the second position information corresponding to the wire passage hole in the first image are obtained.

[0148] The second photo is input into a pre-trained target detection neural network model for detection, and the third position information corresponding to the spindle in the second photo and the fourth position information corresponding to the wire passage hole in the second photo are obtained.

[0149] Based on the first and second position information, calculate the first angle between the line connecting the wire hole and the spindle axis in the first photograph and the preset coordinate axis; based on the third and fourth position information, calculate the second angle between the line connecting the wire hole and the spindle axis in the second photograph and the preset coordinate axis.

[0150] The first angle difference is calculated based on the first included angle and the second included angle; the first time difference between the first and second photos is calculated based on the first shooting frequency; and the first actual rotational speed of the spindle is calculated based on the first angle difference and the first time difference.

[0151] In one embodiment, obtaining the set rotational speed of the spindle of the doubling machine and calculating the first shooting frequency based on the set rotational speed includes:

[0152] Obtain the set speed of the spindle of the doubling twister ;

[0153] Obtain the pre-set first undersampling parameters ;

[0154] According to the set speed and the first undersampling parameter Calculate the first shooting frequency .

[0155] In one embodiment, the step of inputting the first photograph into a pre-trained target detection neural network model for detection, and obtaining the first position information corresponding to the spindle in the first photograph and the second position information corresponding to the wire passage hole in the first photograph, includes:

[0156] The first photo is input into a pre-trained target detection neural network model for detection, and a first bounding box for defining the spindle in the first photo and a second bounding box for defining the thread hole in the first photo are obtained.

[0157] Get the first horizontal coordinates of the top-left vertex of the first bounding box. and the first lateral length of the first bounding box ; set the first horizontal coordinate and the first horizontal length Record this as the first position information;

[0158] Get the second horizontal coordinates corresponding to the top left vertex of the second bounding box. and the second lateral length of the second bounding box ; set the second horizontal coordinate and the second lateral length This is recorded as the second position information.

[0159] In one embodiment, calculating the first angle between the line connecting the wire guide hole and the spindle axis in the first photograph and a preset coordinate axis based on the first position information and the second position information includes:

[0160] Get the radius of the spindle ;

[0161] Obtain the fifth x-coordinate of the spindle's axis in the first image. :

[0162] ;

[0163] Obtain the sixth x-coordinate of the center point of the thread hole in the first image. :

[0164] ;

[0165] Calculate the first horizontal distance between the fifth and sixth horizontal coordinates. :

[0166] ;

[0167] According to the radius and the first lateral distance Calculate the first included angle :

[0168] .

[0169] In one embodiment, the step of inputting the second photograph into a pre-trained target detection neural network model for detection, and obtaining the third position information corresponding to the spindle in the second photograph and the fourth position information corresponding to the wire passage hole in the second photograph, includes:

[0170] The second photo is input into a pre-trained target detection neural network model for detection, and a third bounding box for defining the spindle in the second photo and a fourth bounding box for defining the thread hole in the second photo are obtained accordingly.

[0171] Get the third horizontal coordinates of the top-left vertex of the third bounding box. And the third horizontal length of the third bounding box ; the third horizontal coordinate and the third horizontal length Recorded as third position information;

[0172] Get the fourth horizontal coordinate of the top-left vertex of the fourth bounding box. And the fourth horizontal length of the fourth bounding box ; the fourth horizontal coordinate and the fourth horizontal length This is recorded as the fourth position information.

[0173] In one embodiment, calculating the second angle between the line connecting the wire guide hole and the spindle axis in the second photograph and a preset coordinate axis based on the third and fourth position information includes:

[0174] Get the radius of the spindle ;

[0175] Obtain the seventh x-coordinate of the spindle's axis in the second image. :

[0176] ;

[0177] Obtain the eighth x-coordinate of the center point of the thread hole in the second photo. :

[0178] ;

[0179] Calculate the seventh x-coordinate With the eighth x-axis The second lateral distance between :

[0180] ;

[0181] According to the radius and the second lateral distance Calculate the second included angle :

[0182] .

[0183] In one embodiment, the step of calculating the first angle difference based on the first included angle and the second included angle, calculating the first time difference between the first and second photos based on the first shooting frequency, and calculating the first actual rotational speed of the spindle based on the first angle difference and the first time difference includes:

[0184] According to the first included angle Second angle Calculate the first angle difference ;

[0185] According to the first shooting frequency The first time difference between the first and second photos is calculated accordingly. ;

[0186] Based on the first angle difference and the first time difference Calculate the first actual rotational speed of the spindle. .

[0187] In one embodiment, the method further includes: obtaining a pre-set second undersampling parameter. ;

[0188] According to the set speed and the second undersampling parameter Calculate the second shooting frequency :

[0189] ;

[0190] According to the second shooting frequency Take a third and fourth photograph of the same ingot;

[0191] The third photo is input into a pre-trained target detection neural network model for detection, and the fifth position information corresponding to the spindle in the third photo and the sixth position information corresponding to the wire passage hole in the third photo are obtained.

[0192] The fourth photo is input into a pre-trained target detection neural network model for detection, and the seventh position information corresponding to the spindle in the fourth photo and the eighth position information corresponding to the wire hole in the fourth photo are obtained.

[0193] Based on the fifth and sixth position information, calculate the third included angle between the line connecting the wire hole and the spindle axis in the third photograph and the preset coordinate axis. Based on the seventh and eighth position information, calculate the fourth included angle between the line connecting the wire guide hole and the spindle axis in the fourth photograph and the preset coordinate axis. ;

[0194] According to the third included angle and the fourth angle Calculate the second angle difference ;

[0195] According to the second shooting frequency The second time difference between the first and second photos is calculated accordingly. ;

[0196] Based on the second angle difference and the second time difference Calculate the second actual rotational speed of the spindle. ;

[0197] Determine the first actual speed Is it equal to the second actual speed? If so, then the first actual rotational speed will be... Or the second actual speed Record as actual speed .

[0198] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0199] Example 2

[0200] Furthermore, such as Figure 10 , 11 As shown, it includes: an AGV base, a column, a camera for photographing the spindle, a stroboscope for lighting, a battery, and the aforementioned computer equipment; the bottom of the column is fixedly connected to the AGV base; the battery is fixedly connected to the AGV base; the stroboscope and the camera are both fixedly connected to the side wall of the column; the battery is electrically connected to the camera, the stroboscope, the AGV base, and the computer equipment respectively.

[0201] By mounting an AI camera and flash on a mobile AGV chassis, and having the AGV continuously patrol and inspect the designated range of machines in the workshop with the spindle speed detection equipment, one robot can be used for automatic spindle speed detection of a large number of machines.

[0202] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0203] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A vision-based spindie measurement method, characterized by, The method comprises: acquiring a set rotating speed of a spindle of a double twister, and calculating a first shooting frequency according to the set rotating speed; shooting a first photo and a second photo of the same spindle according to the first shooting frequency; inputting the first photo into a pre-trained target detection neural network model for detection to obtain first position information of the spindle in the first photo and second position information of a wire passing hole in the first photo; inputting the second photo into the pre-trained target detection neural network model for detection to obtain third position information of the spindle in the second photo and fourth position information of the wire passing hole in the second photo; calculating a first included angle between a connecting line of the wire passing hole and the spindle axis in the first photo and a preset coordinate axis according to the first position information and the second position information, and calculating a second included angle between a connecting line of the wire passing hole and the spindle axis in the second photo and the preset coordinate axis according to the third position information and the fourth position information; calculating a first angle difference value according to the first included angle and the second included angle, and calculating a first time difference value between the first photo and the second photo according to the first shooting frequency; calculating a first actual rotating speed of the spindle according to the first angle difference value and the first time difference value; The method comprises: Acquiring a set rotational speed of spindles of a double twisting machine ; acquiring a first pre-set under-sampling parameter ; According to the set rotation speed and the first undersampling parameter , a first shooting frequency is calculated.

2. A vision-based spindown measurement method according to claim 1, characterized in that, The method comprises: inputting the first photo into the pre-trained target detection neural network model for detection to obtain a first bounding box for framing the spindle in the first photo and a second bounding box for framing the wire passing hole in the first photo; obtaining a first horizontal coordinate corresponding to a top-left vertex of the first bounding box and a first horizontal length of the first bounding box ; recording the first horizontal coordinate and the first horizontal length as first position information; acquire a second horizontal coordinate corresponding to a top-left vertex of the second bounding box and a second horizontal length of the second bounding box ; record the second horizontal coordinate and the second horizontal length as the second position information.

3. A vision-based spindown measurement method according to claim 2, wherein, The method comprises: acquiring the radius of the spool ; acquiring a fifth horizontal coordinate where the shaft center of the spindle in the first photo is located : ； acquiring a sixth horizontal coordinate where a center point of the threading hole in the first photograph is located : ； calculating a first lateral distance between the fifth lateral coordinate and the sixth lateral coordinate : ； According to the radius and the first lateral distance the first included angle is calculated 。 4. A vision-based spindown measurement method according to claim 3, wherein, The method comprises: inputting the second photo into the pre-trained target detection neural network model for detection to obtain a third bounding box for framing the spindle in the second photo and a fourth bounding box for framing the wire passing hole in the second photo; a third horizontal coordinate corresponding to a top-left corner vertex of the third bounding box and a third horizontal length of the third bounding box ; the third horizontal coordinate and the third horizontal length are denoted as third position information; a fourth horizontal coordinate corresponding to a top-left vertex of the fourth bounding box and a fourth horizontal length of the fourth bounding box ; the fourth horizontal coordinate and the fourth horizontal length are denoted as fourth position information.

5. A vision-based spindown measurement method according to claim 4, wherein, The method comprises: acquiring the radius of the spool ; acquiring a seventh abscissa of the axis of the spindle in the second photograph : ； acquiring an eighth horizontal coordinate where a center point of the threading hole in the second photo is located ; ； Calculate the seventh x-coordinate With the eighth horizontal coordinate The second lateral distance between : ； According to the radius and the second lateral distance the second included angle is calculated 。 6. A vision-based spindown measurement method according to claim 5, wherein, The method comprises: calculating a first angle difference value according to the first included angle and the second included angle, and calculating a first time difference value between the first photo and the second photo according to the first shooting frequency; According to the first included angle and the second included angle calculating the first angle difference : ; According to the first shooting frequency corresponding to calculating a first time difference value between the first photo and the second photo ； According to the first angle difference and the first time difference , a first actual rotation speed of the spool is calculated 。 7. A vision-based spindown measurement method according to claim 6, wherein, calculating a first actual rotating speed of the spindle according to the first angle difference value and the first time difference value; The method further comprises: acquiring a pre-set second under-sampling parameter ; According to the set rotation speed and the second undersampling parameter , a second imaging frequency is calculated. ； in accordance with the second photographing frequency a third photograph and a fourth photograph are taken of the same spindle The third photo is input into a pre-trained target detection neural network model for detection to obtain fifth position information corresponding to the ingot in the third photo and sixth position information corresponding to the wire hole in the third photo; The fourth photo is input into a pre-trained target detection neural network model for detection to obtain seventh position information corresponding to the ingot in the fourth photo and eighth position information corresponding to the wire hole in the fourth photo; According to the fifth position information and the sixth position information, a third included angle between a connecting line of the wire hole and the spindle axis in the third photo and a preset coordinate axis is calculated ; according to the seventh position information and the eighth position information, a fourth included angle between a connecting line of the wire hole and the spindle axis in the fourth photo and the preset coordinate axis is calculated ; According to the third included angle and the fourth included angle Calculate the second angle difference ; According to the second shooting frequency corresponding to calculating a second time difference value between the first photo and the second photo ； According to the second angle difference and the second time difference , a second actual rotation speed of the spool is calculated ； determining whether the first actual rotational speed is equal to the second actual rotational speed , and if so, setting the first actual rotational speed or the second actual rotational speed as the actual rotational speed .

8. A computer device, comprising: The computer device comprises a memory and a processor; the memory is connected with the processor; the memory is used for storing a computer program; and the processor is used for calling the computer program to enable the computer device to execute the method in any one of claims 1 to 7.

9. A vision-based spindown measurement robot, characterized by The computer device comprises: an AGV base, a stand, a camera for shooting an ingot, a stroboscope for lighting, a battery, and the computer device according to claim 8; the bottom of the stand is fixedly connected with the AGV base; and the battery is fixedly connected with the AGV base; the stroboscope and the camera are both fixedly connected with the side wall of the stand; the battery is electrically connected with the camera, the stroboscope, the AGV base, and the computer device respectively.

Images