Casing pipe residual buckle identification method, system and device, medium and product

Abstract

The invention discloses a sleeve residual buckle identification method, system and device, a medium and a product, and belongs to the technical field of computer vision, and the method comprises the steps: obtaining the size value of a wellhead label in the visual field of a camera and the distance between the camera and the wellhead; according to the size value of the wellhead label and the distance between the camera and the wellhead, the optimal focal length adjusting multiplying power of the camera is determined; focusing the camera according to the optimal focusing magnification of the camera; acquiring a sleeve residual buckle image by using the focused camera; identifying residual buckles in the sleeve residual buckle image to obtain a residual buckle identification result; and according to the residual buckle identification result, judging whether the residual buckle of the sleeve is qualified or not. According to the invention, accurate identification of the residual buckle of the casing pipe is realized, and the technical problem that the residual buckle of the casing pipe cannot be automatically and accurately identified at present is solved.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and in particular to a casing residual buckle recognition method, system, equipment, medium and product. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] In oilfield operations, the quality of residual casing thread has always been a critical factor in ensuring the safety and quality of casing running. Traditional methods for determining residual casing thread quality rely primarily on manual observation and simple methods. These methods are inefficient, susceptible to subjective factors, and difficult to monitor comprehensively, accurately, and in a timely manner. This can lead to misjudgments and missed detections in actual production, impacting production progress and quality, and even posing safety risks.

[0004] In the oilfield production sector, a technology solution exists that uses cameras installed on the drill floor to capture images of the drilling area and monitor drilling operations based on these images. However, when capturing images, the drill floor camera does not perform adaptive focusing based on the distance between the drill floor camera and the wellhead. This fails to ensure image clarity and, consequently, cannot accurately identify remaining casing threads. Summary of the Invention

[0005] In order to solve the above problems, the present invention proposes a casing residual buckle identification method, system, equipment, medium and product, which realize accurate identification of casing residual buckles.

[0006] To achieve the above object, the present invention adopts the following technical solutions: In the first aspect, a casing residual buckle identification method is proposed, comprising: Get the wellhead label size value in the camera's field of view and the distance between the camera and the wellhead; Determine the optimal zoom ratio of the camera based on the wellhead label size and the distance between the camera and the wellhead; Adjust the focus of the camera according to its optimal focus magnification; Use the focused camera to obtain the image of the remaining buckle of the casing; Identify the remaining buckles in the casing remaining buckle image to obtain a remaining buckle identification result; According to the residual thread identification result, determine whether the residual thread of the casing is qualified.

[0007] Further, it is determined whether the residual buckle recognition result is within the set residual buckle qualified range; When the residual buckle identification result is within the qualified residual buckle range, the casing residual buckle is determined to be qualified; When the residual buckle identification result is not within the qualified residual buckle range, it is determined that the casing residual buckle is abnormal.

[0008] Furthermore, an edge detection method is used to identify the excess buckles in the casing excess buckle image.

[0009] Furthermore, the optimal focus ratio of the camera is determined based on the size of the wellhead label in the camera's field of view, the distance between the camera and the wellhead, and a trained focus ratio recognition model; wherein the focus ratio recognition model is constructed using a wavelet neural network.

[0010] Furthermore, the optimal performance indicator of the model refers to the minimum mean square error between the predicted output of the focal length magnification recognition model and the true label.

[0011] Furthermore, the parameters used in the focus ratio recognition model are optimal parameters; the optimal parameters are based on the optimal performance indicators of the model, and the focus ratio recognition model is solved by using the wellhead label size value in the camera field of view with known optimal focus ratio and the distance between the camera and the wellhead.

[0012] Furthermore, the particle swarm algorithm is used to determine the optimal parameters of the focus adjustment magnification recognition model.

[0013] Furthermore, a wellhead label image is obtained; and based on the wellhead label image, a size value of the wellhead label in the camera field of view is determined.

[0014] Secondly, a casing residual buckle identification system is proposed, comprising: A data acquisition unit is used to obtain the size value of the wellhead label in the camera's field of view and the distance between the camera and the wellhead; The automatic focusing module is used to determine the optimal focus magnification of the camera based on the size of the wellhead label and the distance between the camera and the wellhead; and focus the camera according to the optimal focus magnification of the camera; The camera is used to obtain an image of the remaining buckle of the casing after focusing to an optimal focal length magnification; A residual buckle recognition module is used to identify the residual buckles in the casing residual buckle image and obtain a residual buckle recognition result; The determination storage module is used to determine whether the remaining buckles of the casing are qualified according to the remaining buckle identification result.

[0015] In a third aspect, a computer device is provided, comprising: a processor adapted to execute a computer program; A computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the method for identifying remaining casing buckles proposed in the first aspect is implemented.

[0016] In a fourth aspect, a computer-readable storage medium is proposed, wherein the computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor and executing a casing residual buckle identification method proposed in the first aspect.

[0017] In a fifth aspect, a computer program product is proposed, which includes a computer program. When the computer program is executed by a processor, it implements the method for identifying remaining buckles of casing proposed in the first aspect.

[0018] Compared with the prior art, the present invention has the following beneficial effects: The present invention proposes a method, system, equipment, medium and product for identifying casing residual buckles. The method first determines the size value of the wellhead label in the camera's field of view and the distance between the camera and the wellhead, and determines the optimal focusing magnification of the camera based on the size value of the wellhead label in the camera's field of view and the distance between the camera and the wellhead; then, the drilling platform camera is focused according to the optimal focusing magnification, so that the focused camera can obtain the clearest casing residual buckle image; when the casing residual buckle image is used to identify the casing residual buckle, the accuracy of the casing residual buckle identification is guaranteed.

[0019] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The drawings in the specification, which constitute a part of this application, are used to provide further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute improper limitations on this application.

[0021] Figure 1 A flow chart of a method for identifying excess casing buckles disclosed in an embodiment; Figure 2 A block diagram of a casing remaining buckle identification system disclosed in an embodiment; Figure 3 The image of the remaining buckle of the sleeve disclosed in the embodiment; Figure 4 This is the result of identifying the remaining buckles of the casing disclosed in the embodiment. DETAILED DESCRIPTION

[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0023] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present application belongs.

[0024] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0025] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0026] Example 1 In order to achieve accurate identification of casing surplus buckles, in this embodiment, a casing surplus buckle identification method is disclosed, which can automatically identify casing surplus buckles and then determine whether the quality of casing surplus buckles is qualified, reduce dependence on manual experience, and effectively avoid operational risks and accidents caused by casing surplus buckle problems.

[0027] This embodiment discloses a method for identifying excess buckles of casing. Figure 1 Shown, including: Get the wellhead label size value in the camera's field of view and the distance between the camera and the wellhead; Determine the optimal zoom ratio of the camera based on the wellhead label size and the distance between the camera and the wellhead; Adjust the focus of the camera according to its optimal focus magnification; Use the focused camera to obtain the image of the remaining buckle of the casing; Identify the remaining buckles in the casing remaining buckle image to obtain a remaining buckle identification result; According to the residual thread identification result, determine whether the residual thread of the casing is qualified.

[0028] This embodiment obtains a wellhead label image; and determines the size value of the wellhead label in the camera field of view based on the wellhead label image.

[0029] In the specific implementation, the trained image detection model is used to detect the labels in the wellhead label image to obtain the label recognition result; the pixels of the label recognition result are calculated to determine the size value of the wellhead label in the camera field of view.

[0030] Before performing specific casing residual buckle identification, this embodiment obtains the wellhead label size value in the camera field of view at different focal lengths in advance, and performs data fitting to obtain the relationship between the label size value and the focal length; thereafter, only knowing the label size, the relationship can be used to obtain the corresponding focal length, and the focal length is the corresponding focus adjustment magnification.

[0031] Preferably, the image detection model is constructed using the YOLO model.

[0032] To accurately determine the optimal focus magnification, this embodiment utilizes the wellhead label size in the camera's field of view, the distance between the camera and the wellhead, and a trained focus magnification recognition model to determine the camera's optimal focus magnification. The focus magnification recognition model uses the wellhead label size and the distance between the camera and the wellhead as inputs and the camera's optimal focus magnification as output, and is constructed using a wavelet neural network (WNN). To further ensure the optimal focus magnification is determined, the focus magnification recognition model uses optimal parameters. These optimal parameters are obtained by solving the focus magnification recognition model using the wellhead label size in the camera's field of view when the optimal focus magnification is known.

[0033] The focus ratio recognition model uses a wavelet layer to extract the input wellhead label size value and the distance between the camera and the wellhead. The fully connected layer is used to fully connect the features output by the wavelet layer. Then, the output layer calculates the distance value and the optimal focus ratio based on the output of the fully connected layer and outputs them.

[0034] The wavelet layer uses a wavelet function as an activation function to transform the wavelet basis coefficients to extract features from the input data. The fully connected layer receives the output of the wavelet layer and performs a linear transformation using weights and biases to further extract and combine features. The output layer contains two independent neurons, one for the distance value and the other for the optimal focus magnification.

[0035] The reasoning process of the focus adjustment magnification recognition model constructed in this embodiment is based on the pinhole imaging formula, that is, 1 / u + 1 / v = 1 / f, where u is the object distance, that is, the distance between the camera and the wellhead; v is the image distance, which is related to the size of the wellhead label in the camera's field of view; and f is the focal length, that is, the focus adjustment magnification.

[0036] The focus ratio is determined from distance and tag size data. Based on the principle of pinhole imaging, the optimal focus ratio (i.e., focal length f) can be calculated from the object distance u and the image distance v. In the focus ratio recognition model, the image distance v is related to the size of the wellhead tag in the camera's field of view, as the size of the tag in the image reflects its projected size on the photosensitive element. Therefore, the focus ratio recognition model uses a regression algorithm to learn the relationship between the distance u and the wellhead tag size in the camera's field of view (which indirectly reflects v) to predict the optimal focus ratio f.

[0037] The input and output relationship of the focus magnification recognition model is as follows: the input is the distance u between the camera and the wellhead and the size of the wellhead label in the camera's field of view (related to the image distance v), and the output is the optimal focus magnification f. The model establishes a mapping from input to output by learning the physical relationship between u, v, and f or the statistical patterns in the data.

[0038] The trained focus ratio recognition model takes as input the new camera-to-wellhead distance and the wellhead label size in the camera's field of view. It extracts features through convolutional and pooling layers, and finally outputs the predicted optimal focus ratio through a fully connected layer. This reasoning process is based on the pinhole imaging formula: 1 / u + 1 / v = 1 / f, where u is the object distance, v is the image distance, and f is the focal length. The model approximates this formula by learning patterns from the training data.

[0039] The optimal performance indicator of the model refers to the minimum mean square error between the predicted output of the focal length magnification recognition model and the true label.

[0040] In order to improve the speed and accuracy of solving the focus adjustment magnification recognition model, this embodiment adopts a particle swarm optimization algorithm (QPSO algorithm) to determine the optimal parameters of the focus adjustment magnification recognition model.

[0041] After determining the optimal parameters of the focus adjustment magnification model, this embodiment substitutes the optimal parameters into the focus adjustment magnification recognition model. Then, the focus adjustment magnification recognition model using the optimal parameters is trained using the wellhead label size value in the camera field of view with the known optimal focus adjustment magnification and the distance between the camera and the wellhead. After the training is completed, a trained focus adjustment magnification recognition model is obtained.

[0042] The training data used when training the focus magnification recognition model with optimal parameters includes a large number of wellhead label size values and the distance between the camera and the wellhead in the camera field of view that have been marked with the optimal focus magnification.

[0043] The embodiment of the present application collects training data from 146 well teams and performs regression analysis on the collected training data. It summarizes the appropriate value of the distance between the camera and the wellhead and the optimal focus adjustment ratio when the wellhead label size in the camera field of view of real-time target detection is a certain value. This relationship is used to expand the training data to 3,000 samples.

[0044] The embodiment of the present application uses a database table to store training data.

[0045] The specific process of obtaining a trained focus adjustment magnification recognition model includes: (101) Initialize WNN (wavelet neural network) parameters, randomly generate initial wavelet basis coefficients and initialize the weights and biases in the network; (102) Initialize the parameters of the PSO (particle swarm optimization) algorithm, set the number of particles to 30, set the maximum number of iterations to 100, and set the inertia weight or other relevant parameters in the PSO; (103) Preprocess the training data and use the training data to calculate the output of the focus adjustment magnification recognition model for each particle. The mean square error between the predicted output of the calculation model and the true label is used as the fitness value; (104) Based on the fitness value, update the position and velocity of each particle to move it towards a better position. Use the PSO formula to update the position and velocity, including considering the individual best position and the global best position; (105) Repeat steps (103) and (104) until the maximum number of iterations is reached and the optimal parameters are obtained; (106) The training data is used to retrain the focus adjustment magnification recognition model using the optimal parameters. After the training is completed, a trained focus adjustment magnification recognition model is obtained.

[0046] After determining the optimal focus magnification of the camera, this embodiment uses the 3D positioning method in the camera software development kit (SDK) to control the camera to focus on the specified position with the wellhead as the center according to the optimal focus magnification.

[0047] The video stream data of the oilfield operation site is collected using a camera focused to the optimal focal length magnification, and frame processing is performed to obtain image data; the target video frame is selected according to the casing production business specifications; the target video frame is the casing residual buckle image.

[0048] This example configures the camera connection parameters, IP address, and port number according to the Hikvision SDK documentation, initializes the SDK, and establishes a communication connection with the camera.

[0049] The SDK provides an API to obtain the current focal length setting of the camera. Based on the optimal focal length ratio inferred by the model, the SDK API is used to set the optimal focal length ratio of the camera. The camera is then controlled to focus on the specified position centered on the wellhead label.

[0050] In this embodiment, the video stream data of the oilfield operation site is intercepted according to a fixed time size, and then framed to obtain an image sequence; According to the casing production business specifications, when the casing clamp is identified as being directly above the wellhead, it indicates that the make-up operation has begun, and when the casing clamp leaves the vicinity of the wellhead, it indicates that the make-up operation has ended. The image sequence is filtered and the video frames after the make-up is completed are retained as the image sequence of the casing remaining buckle to be identified for subsequent casing remaining buckle identification.

[0051] In this embodiment, an edge detection method is used to identify the remaining threads in the casing remaining thread image, and the remaining thread recognition result is obtained.

[0052] Judge whether the remaining thread recognition result is within the set qualified range of the remaining threads; When the remaining thread recognition result is within the qualified range of the remaining threads, it is determined that the casing remaining threads are qualified; When the remaining thread recognition result is not within the qualified range of the remaining threads, it is determined that the casing remaining threads are abnormal.

[0053] In this embodiment, by performing operations such as region of interest selection, edge detection, smoothing, line detection, and screening and filtering on the casing remaining thread image, the remaining thread recognition result and the remaining thread detection viewable image are obtained; the remaining thread recognition result is compared with the set remaining thread threshold to determine whether the remaining thread recognition result R is within the set qualified range of the remaining threads.

[0054] Among them, the remaining thread threshold includes threshold R0 and threshold R1, and R1 > R0; when R > R1 or R < R0, it means that the remaining thread recognition result is not within the qualified range of the remaining threads. At this time, it is determined that the casing remaining threads are abnormal; when R0 ≤ R ≤ R1, it means that the remaining thread recognition result is within the qualified range of the remaining threads. At this time, it is determined that the casing remaining threads are qualified.

[0055] When it is determined that the casing remaining threads are abnormal, an alarm message is sent, and the real-time analysis result of determining that the casing remaining threads are abnormal is stored and visualized for the supervisors to view and process.

[0056] By identifying the casing remaining threads obtained in the actual operation, the effectiveness of a casing remaining thread identification method disclosed in this embodiment is verified. The casing remaining thread image obtained by using the casing remaining thread identification method disclosed in this embodiment is as Figure 3 shown. After that, using the casing remaining thread identification method disclosed in this embodiment, the Figure 3 image shown is identified, and it is determined that the casing remaining thread recognition result is 3 threads, as shown by the red line segment in Figure 4 . Finally, it is determined that this make-up does not conform to the make-up specification, and the casing remaining thread recognition result is consistent with the actual casing remaining threads, verifying the accuracy of the casing remaining thread identification method disclosed in this embodiment.

[0057] A casing remaining thread identification method disclosed in this embodiment can automatically and real-time detect the state of the casing remaining threads by using the images obtained by the relevant cameras. Compared with the traditional method of relying on manual checking one by one, it is more accurate and efficient, greatly saving a large amount of labor and material costs.

[0058] In terms of production efficiency, the present invention significantly reduces time costs. Furthermore, personnel can access a database to keep abreast of the status of excess casing threads throughout the entire work site. Furthermore, they can directly access data from the database for routine monitoring. On-site alarm equipment can also promptly sound an alarm when an abnormality in excess casing threads is detected, alerting on-site personnel. This significantly reduces potential risks during production operations and effectively improves production efficiency. In terms of application costs, no additional hardware devices such as sensors are required, thus reducing hardware costs. This invention utilizes technologies such as OpenCV and existing cameras at the production site. Simply adding a server and installing the corresponding software on the service end facilitates rapid intelligent recognition of excess casing threads at the lifting operation site.

[0059] Example 2 In this embodiment, a casing remaining buckle identification system is disclosed, such as Figure 2 Shown, including: A data acquisition unit is used to obtain the size value of the wellhead label in the camera's field of view and the distance between the camera and the wellhead; The automatic focusing module is used to determine the optimal focus magnification of the camera based on the size of the wellhead label and the distance between the camera and the wellhead; and focus the camera according to the optimal focus magnification of the camera; The camera is used to obtain an image of the remaining buckle of the casing after focusing to an optimal focal length magnification; A residual buckle recognition module is used to identify the residual buckles in the casing residual buckle image and obtain a residual buckle recognition result; The determination storage module is used to determine whether the remaining buckles of the casing are qualified according to the remaining buckle identification result.

[0060] The present invention also discloses a computer device, comprising: a processor adapted to execute a computer program; A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the method for identifying remaining casing buckles disclosed in embodiment 1 is implemented.

[0061] The present invention also discloses a computer-readable storage medium, which stores a computer program. The computer program is suitable for being loaded by a processor and executing a casing remaining buckle identification method disclosed in Example 1.

[0062] The present invention also discloses a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the method for identifying remaining casing buckles disclosed in Example 1.

[0063] The method disclosed in Example 1 can be directly implemented as a hardware processor, or can be implemented using a combination of hardware and software modules in the processor. The software module can be located in a storage medium mature in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, it will not be described in detail here.

[0064] Those skilled in the art will appreciate that the units and algorithm steps of the various examples described in conjunction with this embodiment can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0065] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A method for identifying excess buckles in casing, characterized in that: include: Get the wellhead label size value in the camera's field of view and the distance between the camera and the wellhead; Determine the optimal zoom ratio of the camera based on the wellhead label size and the distance between the camera and the wellhead; Adjust the focus of the camera according to its optimal focus magnification; Use the focused camera to obtain the image of the remaining buckle of the casing; Identify the remaining buckles in the casing remaining buckle image to obtain a remaining buckle identification result; According to the residual thread identification result, determine whether the residual thread of the casing is qualified.

2. A method for identifying excess casing buckles according to claim 1, characterized in that: Determine whether the residual deduction recognition result is within the set residual deduction qualified range; When the residual buckle identification result is within the qualified residual buckle range, the casing residual buckle is determined to be qualified; When the residual buckle identification result is not within the qualified residual buckle range, it is determined that the casing residual buckle is abnormal.

3. A method for identifying excess casing buckles according to claim 1, characterized in that: The optimal focus ratio of the camera is determined based on the size of the wellhead label in the camera's field of view, the distance between the camera and the wellhead, and a trained focus ratio recognition model. The focus ratio recognition model is constructed using a wavelet neural network.

4. A method for identifying excess casing buckles according to claim 3, characterized in that: The parameters used in the focus ratio recognition model are the optimal parameters. These optimal parameters are aimed at optimizing the model's performance indicators. The focus ratio recognition model is solved by using the wellhead label size in the camera's field of view and the distance between the camera and the wellhead when the optimal focus ratio is known.

5. A method for identifying excess casing buckles according to claim 3, characterized in that: The optimal performance indicator of the model refers to the minimum mean square error between the predicted output of the focal length magnification recognition model and the true label.

6. A method for identifying excess casing buckles according to claim 3, characterized in that: The particle swarm algorithm is used to determine the optimal parameters of the focus adjustment magnification recognition model.

7. A casing remaining buckle identification system, characterized in that: include: A data acquisition unit is used to obtain the size value of the wellhead label in the camera's field of view and the distance between the camera and the wellhead; The automatic focusing module is used to determine the optimal focus magnification of the camera based on the size of the wellhead label and the distance between the camera and the wellhead; and focus the camera according to the optimal focus magnification of the camera; The camera is used to obtain an image of the remaining buckle of the casing after focusing to an optimal focal length magnification; A residual buckle recognition module is used to identify the residual buckles in the casing residual buckle image and obtain a residual buckle recognition result; The determination storage module is used to determine whether the remaining buckles of the casing are qualified according to the remaining buckle identification result.

8. An electronic device, characterized in that: The device comprises: a processor adapted to execute a computer program; A computer-readable storage medium having a computer program stored therein, wherein when the computer program is executed by the processor, the method for identifying a remaining casing buckle according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor and executing the casing remaining buckle identification method according to any one of claims 1 to 6.

10. A computer program product, characterized in that The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the method for identifying remaining casing buckles according to any one of claims 1 to 6.

Images