An array - type connector missing detection method and system

Abstract

The present invention relates to a method and system for detecting the absence of array-type connectors. The detection method is as follows: Based on a preset coding dictionary, generate codes corresponding to the connector arrays in each area; set the codes at visible positions on or near the connector arrays in the form of carriers or direct markings; obtain detection images of each connector array and the corresponding codes through a monitoring workstation provided near the connector arrays; solve the codes in the detection images after preprocessing to obtain the preset arrangement information and quantity information of each connector array; compare the current arrangement information and quantity information with the preset arrangement information and quantity information to obtain the result of the absence information. Compared with the prior art, the present invention realizes the accurate acquisition of connector pictures and corresponding information by a mobile or fixed camera platform, and realizes the rapid extraction of the absence positions, making the absence detection of connectors intelligent and automated.

Description

Technical Field

[0001] The present invention relates to a connector detection system, and more particularly to a method and system for detecting the absence of array connectors. Background Art

[0002] Array connectors are often used for connecting components. Common array connectors are mostly used in steel truss bridges, steel structure buildings, and mechanical structures. Due to the large number of connectors and complex components, the detection of their absence faces multiple challenges. First, manual detection is inefficient and has a low detection frequency. Second, the positions of the connectors are not easily accessible. Third, there is a problem of standardized recording, and it is difficult to standardize the recording and description of the number and form of the missing connectors.

[0003] CN111986161A discloses a method and system for detecting the absence of components. The method includes photographing a target to be detected to obtain an image of the target to be detected; respectively performing component absence detection on the target to be detected in the image of the target to be detected through a deep learning algorithm and an image processing algorithm; and outputting the detection results of the deep learning algorithm and the image processing algorithm.

[0004] The defects of the above technical solutions are as follows: 1. It is impossible to perform specific identification on various array connectors, that is, it is impossible to accurately identify the absence of complex connector arrays. 2. It only quantitatively describes the missing positions after training, and it is impossible to automatically identify the detected components in a new scenario without training. For complex scenarios where it is not easy to have training samples, it lacks usability. 3. The collected images are distorted, and the measured area is inaccurate. Summary of the Invention

[0005] The purpose of the present invention is to overcome the above-mentioned defects existing in the prior art and provide a method and system for detecting the absence of array connectors, which realizes the accurate acquisition of connector pictures and corresponding information by a mobile or fixed camera platform, and realizes the rapid extraction of the missing positions, so that the detection of the absence of connectors is intelligent and automated.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] The first purpose of the present application is to protect a method for detecting the absence of array connectors, including the following steps:

[0008] Generating codes corresponding to the connector arrays in each area based on a preset coding dictionary;

[0009] Correspondingly disposing the codes at visible positions on or near the connector arrays in the form of carriers or direct markings;

[0010] Obtaining detection images of each connector array and the corresponding codes through a monitoring workstation disposed near the connector array;

[0011] Preprocess the acquired detection image;

[0012] Solve the encoding in the preprocessed detection image to obtain the preset arrangement information and quantity information of each connector array;

[0013] Identify each connector in the preprocessed detection image to obtain the current arrangement information and quantity information of each connector array;

[0014] Compare the current arrangement information and quantity information with the preset arrangement information and quantity information to obtain the missing information result.

[0015] Further, the information encoding is one or a combination of two-dimensional barcodes, one-dimensional barcodes, strings, checkerboards, etc.

[0016] Further, the carrier is a rigid or flexible material on which the information encoding can be printed.

[0017] Further, the preprocessing process is: perform perspective transformation on the detection image to obtain a front view of the connector array and the corresponding information encoding, and realize image correction.

[0018] Further, the arrangement information includes the geometric center coordinates of each connector in the connector array;

[0019] The quantity information includes the number of rows and columns of the connector array and the number of connectors corresponding to each row and column.

[0020] Further, in the process of identifying each connector in the preprocessed detection image, the connector is identified by an object detection algorithm based on deep learning or a contour recognition algorithm based on graphics.

[0021] Further, the encoding also includes a structure encoding, and the structure encoding includes calibration information required for perspective transformation, specific parameters and / or thresholds required for an object detection algorithm based on deep learning or a contour recognition algorithm based on graphics.

[0022] Further, in the process of comparing the current arrangement information and quantity information with the preset arrangement information and quantity information, it includes:

[0023] Direct quantity verification, based on the comparison of the number of connectors in each connector array with the number of connectors recorded in the encoding, qualitatively and quantitatively determine the missing state of a specific connector array;

[0024] Dot matrix matching verification: Based on Canny edge detection of the pre - processed detection image or object detection algorithms based on deep learning, calculate the coordinates corresponding to the geometric center of the edge contour, match them with the geometric center coordinates of each connector recorded in the code, obtain the exact missing positions, and achieve accurate verification of the direct quantity verification results.

[0025] Further, the monitoring workstation includes:

[0026] A working track;

[0027] A workbench, which is arranged on the working track and can be displaced along the working track;

[0028] A photographing unit, which is arranged on the workbench and can move to a specific position on the working track and photograph the detection images of each connector array and the corresponding code in a specific posture.

[0029] The second object of the present application is to protect an array - type connector missing detection system, including: a code, a working track, a workbench, a photographing unit, and a background unit. Specifically:

[0030] The code is correspondingly arranged at a visible position on or near the connector array in the form of a carrier or direct marking;

[0031] A working track;

[0032] The workbench is arranged on the working track and can be displaced along the working track;

[0033] The photographing unit is arranged on the workbench and can move to a specific position on the working track and photograph the detection images of each connector array and the corresponding code in a specific posture;

[0034] The background unit is wirelessly or wiredly communicatively connected to the photographing unit. The background unit pre - processes the obtained detection images, decodes the codes in the pre - processed detection images, obtains the preset arrangement information and quantity information of each connector array, identifies each connector in the pre - processed detection images, obtains the current arrangement information and quantity information of each connector array, and finally compares the current arrangement information and quantity information with the preset arrangement information and quantity information to obtain the missing information result.

[0035] Further, the background unit includes a microprocessor, RAM, and ROM, and the microprocessor is electrically connected to the photographing unit.

[0036] Compared with the prior art, the present invention has the following technical advantages:

[0037] 1) It realizes the accurate acquisition of pictures and corresponding information of connecting pieces by a mobile or fixed camera platform, and enables the rapid extraction of missing positions, making the missing detection of connecting pieces intelligent and automated. Specifically, this technical solution uses coding for identification. The advantages are that it can not only provide component information for convenient filing and recording, but on the other hand, the coding system can provide the point coordinates required for perspective correction, perform perspective correction on the image, and thus greatly reduce errors. In addition, for array-type connecting pieces, this method can conveniently locate the missing position of a specific connecting piece and quantitatively describe the missing position coordinates of a single connecting piece.

[0038] 2) In terms of accuracy, by setting identification markers, based on the true size of the identification markers and the actually detected true points in the image, a corresponding relationship can be established between the pixel size and the physical size of the object image, and perspective transformation is performed on the entire image to eliminate the perspective error generated by shooting on the surface of the flange plate. At the same time, the true position of the missing connecting piece can be directly obtained.

[0039] 3) In terms of the detection strategy, by calculating the information of the identification markers, the number information and connecting piece information on the metal surface can be obtained. Combining the results of perspective transformation, on the one hand, the missing quantity and distribution of other structural components on the surface of the flange plate, such as connecting pieces and combined structures, can be obtained automatically; on the other hand, in large-scale scenarios, such as the gusset plates of steel truss bridges, there are thousands of gusset plates, most of which are repetitive, and it is very difficult to perform automatic positioning and recording. Through this technical solution, the size and position information of the gusset plates can be automatically located, the connecting piece numbers can be corresponding one by one, and automatic filing can be performed after detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic flow chart of the method for detecting missing array-type connecting pieces in this technical solution;

[0041] Figure 2 It is a schematic structural diagram of the information coding setting in this technical solution;

[0042] Figure 3 It is a schematic structural diagram of the system for detecting missing array-type connecting pieces in this technical solution.

[0043] In the figure: 0, structural body; 1, connecting piece array; 11, single connecting piece; 2, coding; 3, monitoring workstation; 31, working track; 32, workbench; 33, shooting unit; 34, driving motor; 35, lead screw; 4, working track fixing piece. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0045] Embodiment

[0046] In this embodiment, the method for detecting the absence of array connectors is as follows. Refer to Figure 1 , which includes the following steps:

[0047] Based on a preset coding dictionary, generate the code 2 corresponding to the connector array 1 in each area. The information code 2 is one or a combination of a two-dimensional code, a one-dimensional code, a string, and a checkerboard. Figure 2 shows the form of two-dimensional code encoding and setting, and other methods and combination methods can also be realized.

[0048] Correspondingly set the code 2 in a visible position on or near the connector array 1 in the form of a carrier or direct marking. The carrier is a rigid or flexible material on which the information code 2 can be printed, such as a polymer flexible material or a polymer board, with one side as the printing surface and the other side as the adhesive surface, or a buckle can also be used to fix the above carrier. In a specific application scenario, the connector array 1 can be arranged on any wall surface of the structural body 0. Figure 3 Only taking the lower surface as an example in

[0049] Obtain the detection images of each connector array 1 and the corresponding code 2 through the monitoring workstation 3 arranged near the connector array 1. The monitoring workstation 3 includes a working track 31, a workbench 32, and a photographing unit 33. The workbench 32 is arranged on the working track 31 and can be displaced along the working track 31. The photographing unit 33 is arranged on the workbench 32 and can move to a specific position on the working track 31 and photograph the detection images of each connector array 1 and the corresponding code 2 in a specific posture.

[0050] Preprocess the obtained detection images. The preprocessing process is as follows: perform perspective transformation on the detection images to obtain the front view of the connector array 1 and the corresponding information code 2, and realize the correction of the images. When necessary, the preprocessing process also includes common preprocessing steps such as binarization and grayscale.

[0051] Solve the code 2 in the preprocessed detection images to obtain the preset arrangement information and quantity information of each connector array 1. The arrangement information includes the geometric center coordinates of each connector in the connector array 1; the quantity information includes the number of rows and columns of the connector array 1 and the number of connectors corresponding to each row and column. The code 2 also includes a structure code, and the structure code includes the calibration information required for perspective transformation, specific parameters and / or thresholds required for object detection algorithms based on deep learning or contour recognition algorithms based on graphics. The image perspective transformation involved in this embodiment can be realized by using existing image perspective transformation technologies. The object detection algorithms based on deep learning or contour recognition algorithms described in this technical solution can be selected from existing mature mainstream algorithms. Since the recognition process of array connectors is relatively simple and rapid, simple and fast mainstream algorithms can be selected.

[0052] Identify each connecting piece in the preprocessed detected image to obtain the current arrangement information and quantity information of each connecting piece array 1. During the process of identifying each connecting piece in the preprocessed detected image, identify the connecting piece through an object detection algorithm based on deep learning or a contour recognition algorithm based on graphics.

[0053] Compare the current arrangement information and quantity information with the preset arrangement information and quantity information to obtain the missing information result. During the process of comparing the current arrangement information and quantity information with the preset arrangement information and quantity information, it includes: direct quantity verification, comparing the quantity of connecting pieces in each connecting piece array 1 with the quantity of connecting pieces recorded in the code 2 to qualitatively and quantitatively determine the missing state of a specific connecting piece array 1; dot matrix matching verification, based on performing Canny image edge detection on the preprocessed detected image or an object detection algorithm based on deep learning, calculate the coordinates corresponding to the geometric center of the edge contour, and match them with the geometric center coordinates of each connecting piece recorded in the code 2 to obtain the exact missing position and accurately verify the result of the direct quantity verification. The above Canny image edge detection or object detection algorithm based on deep learning can both adopt the algorithms in the prior art.

[0054] In this embodiment, an array-type connecting piece missing detection system includes: a code 2, a working track 31, a workbench 32, a photographing unit 33, and a background unit. See Figure 2 and Figure 3 。

[0055] The code 2 is correspondingly arranged at a visible position on or near the connecting piece array 1 in the form of a carrier or direct marking. The working track 31, as the moving track of the workbench 32, can be customized according to the scene and area of the detection area, so that the workbench 32 can meet the monitoring and inspection requirements of the workbench 32. Specifically, in implementation, the working track 31 can be a guide chute.

[0056] The workbench 32 is arranged on the working track 31 and can be displaced along the working track 31. Specifically, in implementation, a slider matching the guide chute is provided on the lower surface of the workbench 32 and can be embedded in the guide chute for displacement. In addition, pulleys can also be arranged on the lower surface of the workbench 32, which can also meet the ideal inspection and movement requirements. The driving component of the workbench 32 can be selected as the transmission cooperation of a driving motor 34 and a lead screw 35, so that the lead screw 35 and the workbench 32 form a thread match to realize the linear displacement drive of the workbench 32. In addition, if the form of pulleys is adopted, a driving motor and a steering servo can be matched with the pulleys to realize the curve displacement drive of the workbench 32. In this case, the matching of a curve-shaped working track 31 can be achieved.

[0057] The photographing unit 33 is disposed on the workbench 32 and can move to a specific position on the work track 31 and photograph the detection images of each connector array 1 and the corresponding code 2 in a specific posture. The photographing unit 33 can be selected as a high-resolution industrial camera. Specifically in implementation, it further includes a work track fixing member 4, and the work track 31 is fixed to the lower surface of the structural body 0 through the work track fixing member 4.

[0058] The background unit is wirelessly or wiredly communicatively connected to the photographing unit 33. The background unit preprocesses the acquired detection images, solves the code 2 in the preprocessed detection images, obtains the preset arrangement information and quantity information of each connector array 1, identifies each connector in the preprocessed detection images, obtains the current arrangement information and quantity information of each connector array 1, and finally compares the current arrangement information and quantity information with the preset arrangement information and quantity information to obtain the missing information result. The background unit includes a microprocessor, a RAM, and a ROM. The microprocessor is electrically connected to the photographing unit 33. Among them, the ROM stores executable programs for the above-mentioned preprocessing, solving, identifying the connectors, and comparing the current arrangement quantity information with the preset arrangement quantity information. The microprocessor is selected as an ARM processor or a processor with an x86 architecture. In addition, the background unit further includes a signal transceiver unit, which is electrically connected to the microprocessor. The signal transceiver unit is wirelessly communicatively connected to an external computer terminal, thereby realizing the information transmission to the user end. The signal transceiver unit can be implemented by using an existing wireless signal transceiver module.

[0059] The above description of the embodiments is to enable those of ordinary skill in the art in this technical field to understand and use the invention. It is obvious that those skilled in the art can easily make various modifications to these embodiments and apply the general principles described herein to other embodiments without creative efforts. Therefore, the present invention is not limited to the above embodiments, and all improvements and modifications made by those skilled in the art without departing from the scope of the present invention according to the disclosure of the present invention should be within the protection scope of the present invention.

Claims

1. An array connector missing detection method, characterized in that, Including the following steps: Generating a code (2) corresponding to the connector array (1) of each area based on a preset coding dictionary; Correspondingly setting the code (2) at a visible position on or near the connector array (1) in the form of a carrier or direct marking; Obtaining a detection image of each connector array (1) and the corresponding code (2) through a monitoring workstation (3) provided near the connector array (1); Preprocessing the obtained detection image; Solving the code (2) in the preprocessed detection image to obtain the preset arrangement information and quantity information of each connector array (1); Identifying each connector in the preprocessed detection image to obtain the current arrangement information and quantity information of each connector array (1); Comparing the current arrangement information and quantity information with the preset arrangement information and quantity information to obtain a missing information result; The arrangement information includes the geometric center coordinates of each connector in the connector array (1); The quantity information includes the number of rows and columns of the connector array (1) and the number of connectors corresponding to each row and column; During the process of comparing the current arrangement information and quantity information with the preset arrangement information and quantity information, it includes: Direct quantity verification, comparing the number of connectors in each connector array (1) with the number of connectors recorded in the code (2) to qualitatively and quantitatively determine the missing state of a specific connector array (1); Dot matrix matching verification, based on performing Canny image edge detection on the preprocessed detection image or a target detection algorithm based on deep learning, calculating the coordinates corresponding to the geometric center of the edge contour, and matching them with the geometric center coordinates of each connector recorded in the code (2) to obtain the precise missing position and accurately verify the result of the direct quantity verification.

2. The method for detecting the absence of an array-type connecting member according to claim 1, wherein, The code (2) is one or a combination of a two-dimensional code, a one-dimensional code, a string, a checkerboard, etc.

3. The method for detecting the absence of an array-type connecting member according to claim 1, wherein The preprocessing process is: performing perspective transformation on the detection image to obtain a front view of the connector array (1) and the corresponding information code (2) to correct the image.

4. The method for detecting the absence of an array - type connecting member according to claim 1, wherein, During the process of identifying each connector in the preprocessed detection image, the connector is identified through a target detection algorithm based on deep learning or a contour recognition algorithm based on graphics.

5. The method for detecting the absence of an array-type connector according to claim 1, wherein The code (2) further includes a structure code, and the structure code includes calibration information required for perspective transformation, specific parameters and / or thresholds required for a target detection algorithm based on deep learning or a contour recognition algorithm based on graphics.

6. The method for detecting the absence of an array-type connecting member according to claim 1, characterized in that The monitoring workstation (3) includes: A working track (31); A workbench (32), provided on the working track (31) and displaceable along the working track (31); A photographing unit (33), provided on the workbench (32), capable of moving to a specific position on the working track (31) and photographing the detection image of each connector array (1) and the corresponding code (2) in a specific posture.

7. An array - type connector missing detection system, characterized in that, Including: A code (2), correspondingly set at a visible position on or near the connector array (1) in the form of a carrier or direct marking; A working track (31); A workbench (32) is provided on the work track (31) and can be displaced along the work track (31). A photographing unit (33) is provided on the workbench (32), and can move to a specific position on the work track (31) and photograph detection images of each connector array (1) and the corresponding code (2) in a specific posture. A background unit is wirelessly or wiredly communicatively connected to the photographing unit (33). The background unit preprocesses the acquired detection images, decodes the code (2) in the preprocessed detection images to obtain the preset arrangement information and quantity information of each connector array (1), identifies each connector in the preprocessed detection images to obtain the current arrangement information and quantity information of each connector array (1), and finally compares the current arrangement information and quantity information with the preset arrangement information and quantity information to obtain a missing information result. The specific process includes: Based on a preset code dictionary, generating the code (2) corresponding to each connector array (1) in each area. Correspondingly setting the code (2) at a visible position on or near the connector array (1) in the form of a carrier or direct marking. Obtaining detection images of each connector array (1) and the corresponding code (2) through a monitoring workstation (3) provided near the connector array (1). Preprocessing the acquired detection images. Decoding the code (2) in the preprocessed detection images to obtain the preset arrangement information and quantity information of each connector array (1). Identifying each connector in the preprocessed detection images to obtain the current arrangement information and quantity information of each connector array (1). Comparing the current arrangement information and quantity information with the preset arrangement information and quantity information to obtain a missing information result. The arrangement information includes the geometric center coordinates of each connector in the connector array (1). The quantity information includes the number of rows and columns of the connector array (1) and the number of connectors corresponding to each row and column. During the process of comparing the current arrangement information and quantity information with the preset arrangement information and quantity information, it includes: Direct quantity verification, comparing the number of connectors in each connector array (1) with the number of connectors recorded in the code (2) to qualitatively and quantitatively determine the missing state of a specific connector array (1). Dot matrix matching verification, based on performing Canny image edge detection on the preprocessed detection images or a target detection algorithm based on deep learning, calculating the coordinates corresponding to the geometric center of the edge contour, and matching them with the geometric center coordinates of each connector recorded in the code (2) to obtain the exact missing position and accurately verify the result of the direct quantity verification.

8. An array - type connector missing detection system according to claim 7, wherein, The background unit includes a microprocessor, RAM, and ROM, and the microprocessor is electrically connected to the photographing unit (33).

Images