Optical cable sheath defect identification device and system based on machine vision

Abstract

The invention relates to the technical field of optical cable detection, in particular to an optical cable sheath defect identification device and system based on machine vision, and the device comprises a housing which is internally provided with a first detection space and a second detection space; the visual identification mechanism is arranged in the first detection space and the second detection space and is used for identifying and detecting the sheath of the optical cable, the shell is provided with a through hole penetrating through the first detection space and the second detection space, and the through hole is used for allowing the optical cable to penetrate through. The visual identification mechanism comprises a first visual camera and a second visual camera which are arranged in the first detection space; the visual identification mechanism further comprises a third visual camera and a fourth visual camera which are arranged in the second detection space. The device can realize omnibearing real-time online monitoring, can timely discover and process potential quality problems, prevents defective products from flowing out, improves the product quality and the production efficiency, and reduces the use cost.

Description

Technical Field

[0001] The present invention relates to the field of optical cable detection technology, and in particular to an optical cable sheath defect recognition device and system based on machine vision. Background Art

[0002] As an indispensable infrastructure in modern communication networks, the quality of the sheath of optical cables is directly related to the durability and service life of optical cables. However, during the production and laying of optical cables, various defects such as scratches, stains, and breakage may occur on the sheath of optical cables due to process, material or environmental factors. These defects not only affect the aesthetics of the optical cable, but more importantly, they may affect the transmission performance and safety of the optical cable. At present, the detection of optical cable sheath defects mainly relies on manual visual inspection, which has many shortcomings: low detection efficiency, easy fatigue leading to missed detection and false detection, high labor costs, and difficulty in ensuring the consistency and accuracy of detection. In addition, existing detection methods based on visual deep learning, such as radar and remote sensing technology, are highly accurate, but the detection is time-consuming, pre-processing is complex, and the cost is high, making it difficult to popularize.

[0003] Therefore, a device and system for identifying optical cable sheath defects based on machine vision are proposed. Summary of the invention

[0004] In view of the above problems or problems existing in the prior art, the present invention is proposed.

[0005] Therefore, the object of the present invention is to provide a device and system for identifying optical cable sheath defects based on machine vision.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a device for identifying optical cable sheath defects based on machine vision, comprising:

[0007] A housing having a first detection space and a second detection space disposed therein;

[0008] The visual recognition mechanism is arranged in the first detection space and the second detection space and is used for identifying and detecting the outer sheath of the optical cable.

[0009] As a preferred solution of the optical cable sheath defect identification device based on machine vision of the present invention, wherein: a light source is provided in both the first detection space and the second detection space.

[0010] As a preferred solution of the optical cable sheath defect identification device based on machine vision of the present invention, wherein: the shell is provided with a through hole penetrating the first detection space and the second detection space, and the through hole is used for the optical cable to pass through.

[0011] As a preferred solution of the optical cable sheath defect recognition device based on machine vision of the present invention, wherein: the visual recognition mechanism includes a first visual camera and a second visual camera arranged in the first detection space;

[0012] The visual recognition mechanism further includes a third visual camera and a fourth visual camera disposed in the second detection space;

[0013] The first visual camera, the second visual camera, the third visual camera and the fourth visual camera are respectively used to collect images of the surrounding areas of the optical cable sheath.

[0014] As a preferred solution of the optical cable sheath defect recognition device based on machine vision of the present invention, wherein: the first visual camera and the second visual camera are arranged vertically;

[0015] The third visual camera and the fourth visual camera are arranged vertically.

[0016] As a preferred solution of the optical cable sheath defect identification device based on machine vision of the present invention, wherein: a third detection space is provided in the housing, and the third detection space is provided between the first detection space and the second detection space;

[0017] It also includes a limit support mechanism disposed in the third detection space, the limit support mechanism includes a fixed plate disposed in the third detection space, and a first limit wheel disposed on the fixed plate, the fixed plate is provided with a second limit wheel, and also includes a support wheel disposed on the fixed plate;

[0018] The first limiting wheel, the second limiting wheel and the supporting wheel are distributed in a triangle shape, and the first limiting wheel and the second limiting wheel are arranged on the same side of the optical cable.

[0019] As a preferred solution of the optical cable sheath defect identification device based on machine vision of the present invention, it further includes a guiding mechanism, which includes support plates arranged on both sides of the housing, and a first guide wheel arranged on the support plate, and a second guide wheel is arranged on the support plate;

[0020] The first guide wheel and the second guide wheel are respectively arranged on both sides of the optical cable.

[0021] The beneficial effects of the optical cable sheath defect identification device based on machine vision of the present invention are as follows: the device can realize all-round real-time online monitoring, can timely discover and deal with potential quality problems, avoid the outflow of defective products, improve product quality and production efficiency, and reduce the cost of use.

[0022] In order to solve the above technical problems, the present invention also provides the following technical solutions: a machine vision-based optical cable sheath defect recognition system, comprising the machine vision-based optical cable sheath defect recognition device, and further comprising:

[0023] A position recording module, which includes a speed encoder and a meter recorder;

[0024] A communication module, which is used for data transmission;

[0025] The sound and light alarm module is used to send out different sound and light signals according to the detection results.

[0026] As a preferred solution of the optical cable sheath defect recognition system based on machine vision of the present invention, the speed encoder and meter counter are used to measure the moving speed and length data of the optical cable.

[0027] The beneficial effects of the optical cable sheath defect recognition system based on machine vision of the present invention are as follows: the system has high detection efficiency and high accuracy, and is convenient for locating and alarming defective parts to achieve all-round real-time online monitoring.

[0028] In order to solve the above technical problems, the present invention also provides the following technical solutions: a method for training optical cable defect machine vision, applied to the optical cable sheath defect recognition system based on machine vision, comprising:

[0029] Collect different types of optical cable defect images;

[0030] Build defect target detection model;

[0031] Training the recognition model;

[0032] Build the trained model.

[0033] The beneficial effects of the optical cable defect machine vision training method of the present invention are as follows: machine vision detection can realize automated and continuous detection, greatly improving detection efficiency, reducing labor costs and time costs, while ensuring the stability and consistency of detection results, avoiding manual, time-consuming and labor-intensive work, and being able to automatically record the detection results and location information of the optical cable, making the detection process highly automated, greatly reducing manual intervention, thereby significantly improving detection efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0035] Figure 1 This is a schematic diagram of the overall structure of the optical cable sheath defect recognition device based on machine vision.

[0036] Figure 2 The schematic diagram of the framework of the optical cable sheath defect recognition system based on machine vision.

[0037] Figure 3 Flow chart of the optical cable defect machine vision training method. DETAILED DESCRIPTION

[0038] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0039] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0040] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0041] Example 1, reference Figure 1 , which is the first embodiment of the present invention, provides an optical cable sheath defect recognition device based on machine vision, comprising:

[0042] The housing 100 has a first detection space A and a second detection space C disposed therein;

[0043] The visual recognition mechanism 200 is arranged in the first detection space A and the second detection space C, and is used to identify and detect the outer sheath of the optical cable.

[0044] In this embodiment, the optical cable sheaths in the first detection space A and the second detection space C are identified and detected by the visual recognition mechanism 200 .

[0045] Specifically, light sources are provided in both the first detection space A and the second detection space C. In the present embodiment, the addition of light sources can enhance the detection effect of the visual recognition mechanism 200 .

[0046] Furthermore, the housing 100 is provided with a through hole penetrating the first detection space A and the second detection space C, and the through hole is used for the optical cable to pass through.

[0047] The visual recognition mechanism 200 includes a first visual camera 201 and a second visual camera 202 disposed in the first detection space A;

[0048] The visual recognition mechanism 200 further includes a third visual camera 203 and a fourth visual camera 204 disposed in the second detection space C;

[0049] The first visual camera 201, the second visual camera 202, the third visual camera 203 and the fourth visual camera 204 are respectively used to collect images of the surroundings of the outer sheath of the optical cable.

[0050] Preferably, the first visual camera 201 and the second visual camera 202 are arranged vertically;

[0051] The third visual camera 203 and the fourth visual camera 204 are arranged vertically.

[0052] In this embodiment, after the first visual camera 201 and the second visual camera 202 are vertically arranged, the light source can be set at the center of the angle formed by the first visual camera 201 and the second visual camera 202, and one light source can meet the use of the first visual camera 201 and the second visual camera 202. Similarly, after the third visual camera 203 and the fourth visual camera 204 are vertically arranged, the light source can be set at the center of the angle formed by the third visual camera 203 and the fourth visual camera 204, and one light source can meet the use of the third visual camera 203 and the fourth visual camera 204.

[0053] It should be noted that a third detection space B is provided in the housing 100 , and the third detection space B is provided between the first detection space A and the second detection space C;

[0054] The apparatus further includes a position limiting support mechanism 300 disposed in the third detection space B, the position limiting support mechanism 300 includes a fixed plate 301 disposed in the third detection space B, and a first position limiting wheel 302 disposed on the fixed plate 301, a second position limiting wheel 303 is disposed on the fixed plate 301, and further includes a support wheel 304 disposed on the fixed plate 301;

[0055] The first limiting wheel 302 , the second limiting wheel 303 , and the supporting wheel 304 are distributed in a triangle shape, and the first limiting wheel 302 and the second limiting wheel 303 are arranged on the same side of the optical cable.

[0056] In this embodiment, the first limiting wheel 302, the second limiting wheel 303, and the supporting wheel 304 are all made of rubber material, and are used to fix and stretch the cable, and can also prevent secondary damage to the optical cable.

[0057] The housing 100 further includes a guide mechanism 400, which includes support plates 401 disposed on both sides of the housing 100, and a first guide wheel 402 disposed on the support plate 401, and a second guide wheel 403 disposed on the support plate 401;

[0058] The first guide wheel 402 and the second guide wheel 403 are respectively disposed on both sides of the optical cable.

[0059] In this embodiment, the first guide wheel 402 and the second guide wheel 403 are both made of rubber material, and are used to fix and stretch the cable, and also to prevent secondary damage to the optical cable.

[0060] When in use, the common defect images of the optical cable are photographed and stored through the visual recognition mechanism 200. At least hundreds of pictures are taken for each defect type, such as scratches, stains, breakages, bulges, bubbles and other defect types. After obtaining different types of defect images through the visual recognition mechanism 200, the images are preprocessed to extract the region of interest, and then the region of interest is deep-learned and trained on the official platform. Finally, the defects of the optical cable are judged based on the deep learning features, the defects of the optical cable are identified, and after the recognition model is trained, the trained model is converted into a model format and deployed to the base In the optical cable sheath defect recognition device of machine vision, after the common defect images of the optical cable are photographed and stored, model training is performed on the official platform. The model and a large number of defect type images taken will be stored in the SD card, and the model will be deployed in the machine vision camera. When the machine vision training is completed, it is offline from the computer side. Only the camera needs to be powered to read the previously trained model data from the SD card to realize online detection of defects. When a defect is identified, the machine vision camera will send a signal to the main control, which is convenient for the subsequent encoder to measure the speed and the meter to determine the location of the optical cable defect.

[0061] The optical cable is transported in the through hole, and the first guide wheel 402 and the second guide wheel 403 support and limit the two ends of the optical cable in the device. The first guide wheel 402 and the second guide wheel 403 limit and support the optical cable located in the middle of the device. The first visual camera 201, the second visual camera 202, the third visual camera 203 and the fourth visual camera 204 perform all-round shooting and scanning of the surface of the optical cable to identify defects on the cable surface. The setting of the four visual cameras can prevent mutual identification interference.

[0062] In summary, the device can realize all-round real-time online monitoring, can promptly discover and deal with potential quality problems, avoid the outflow of defective products, improve product quality and production efficiency, and reduce use costs.

[0063] Example 2, reference Figure 1-2, which is the second embodiment of the present invention. Different from the previous embodiment, this embodiment provides a machine vision-based optical cable sheath defect recognition system, including a machine vision-based optical cable sheath defect recognition device, and also including,

[0064] A position recording module, which includes a speed encoder and a meter recorder;

[0065] A communication module, which is used for data transmission;

[0066] The sound and light alarm module is used to send out different sound and light signals according to the detection results.

[0067] Specifically, the speed encoder and meter recorder are used to measure the moving speed and length data of the optical cable.

[0068] In this embodiment, the moving speed and length data of the optical cable can be measured by a speed encoder and a meter, and the distance between the two lenses and the node of the production line can be directly measured. By combining the time and the current detection position, each frame of the detection image can be matched with the position on the optical cable to achieve positioning detection. Once the current image detection result shows that there is a defect, the location of the defect can be directly calculated, which is convenient for later processing.

[0069] The rest of the structure is the same as that of Example 1.

[0070] When in use, the optical cable sheath defect recognition device based on machine vision identifies the surface condition of the optical cable. When a defect occurs, the defect position is recorded by the speed encoder and the meter recorder to locate the defect. The defect data is transmitted and sent through the communication module, and the corresponding type of alarm is issued in conjunction with the sound and light alarm module.

[0071] In summary, the system has high detection efficiency and high accuracy, which facilitates the positioning and alarm of defective parts to achieve all-round real-time online monitoring.

[0072] Example 3, reference Figure 1 to Figure 3 , which is the third embodiment of the present invention. Different from the previous embodiment, this embodiment provides a machine vision training method for optical cable defects, which is applied to an optical cable sheath defect recognition system based on machine vision, including:

[0073] Collect different types of optical cable defect images;

[0074] Build defect target detection model;

[0075] Training the recognition model;

[0076] Build the trained model.

[0077] Specifically, refer to Figure 3First, the K210 camera is used to collect different types of optical cable defect images, and the image is preprocessed to obtain a data set; the k210 chip is used as the core of the system to realize data processing and overall control functions, and the system uses the ov2640 camera to collect stamping images. Secondly, a defect target detection model based on the improved yolov3 is constructed, and the target labels of different types of optical cable sheath defects are annotated using the official training platform, and the data set is sent to the model for training to obtain a defect target area detection model; then, a defect recognition model based on ise-densenet is constructed, and a defect training set is constructed based on defect target detection to train the recognition model; finally, the trained model is converted into a model format, and the K210 development software maixPy IDE is used to deploy the model to the k210 embedded platform. When an optical cable defect appears in the picture, the signal is sent to the host computer and the control end of the computer room through the serial port, and finally the online detection of optical cable defects is realized.

[0078] The rest of the structure is the same as that of Example 2.

[0079] In summary, machine vision inspection can realize automatic and continuous inspection, greatly improving inspection efficiency and reducing labor and time costs. At the same time, it ensures the stability and consistency of the inspection results, avoids manual, time-consuming and labor-intensive work, and can automatically record the inspection results and location information of the optical cable, making the inspection process highly automated, greatly reducing manual intervention, and thus significantly improving inspection efficiency;

[0080] Through the continuous application of artificial intelligence technologies such as deep learning, the machine vision inspection system can automatically learn and understand the various morphological characteristics of optical cable sheath defects, and achieve more intelligent and adaptive detection capabilities. This means that the system can continuously optimize its detection algorithm to meet the needs of optical cable sheath defect detection of different types and specifications.

[0081] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A device for identifying optical cable sheath defects based on machine vision, characterized in that: include, A housing (100) having a first detection space (A) and a second detection space (C) disposed therein; A visual recognition mechanism (200) is arranged in the first detection space (A) and the second detection space (C) and is used to identify and detect the outer sheath of the optical cable.

2. The optical cable sheath defect recognition device based on machine vision according to claim 1, characterized in that: Light sources are provided in both the first detection space (A) and the second detection space (C).

3. The optical cable sheath defect recognition device based on machine vision according to claim 1 or 2, characterized in that: The housing (100) is provided with a through hole penetrating the first detection space (A) and the second detection space (C), and the through hole is used for the optical cable to pass through.

4. The optical cable sheath defect recognition device based on machine vision as claimed in claim 3, characterized in that: The visual recognition mechanism (200) comprises a first visual camera (201) and a second visual camera (202) which are arranged in the first detection space (A); The visual recognition mechanism (200) further comprises a third visual camera (203) and a fourth visual camera (204) arranged in the second detection space (C); The first visual camera (201), the second visual camera (202), the third visual camera (203) and the fourth visual camera (204) are respectively used to collect images of the surroundings of the outer sheath of the optical cable.

5. The optical cable sheath defect recognition device based on machine vision as claimed in claim 4, characterized in that: The first visual camera (201) and the second visual camera (202) are arranged vertically; The third visual camera (203) and the fourth visual camera (204) are arranged vertically.

6. The optical cable sheath defect recognition device based on machine vision as claimed in claim 5, characterized in that: A third detection space (B) is provided in the housing (100), and the third detection space (B) is provided between the first detection space (A) and the second detection space (C); The invention also comprises a position-limiting support mechanism (300) arranged in the third detection space (B), wherein the position-limiting support mechanism (300) comprises a fixing plate (301) arranged in the third detection space (B), and a first position-limiting wheel (302) arranged on the fixing plate (301), a second position-limiting wheel (303) is arranged on the fixing plate (301), and further comprises a support wheel (304) arranged on the fixing plate (301); The first limiting wheel (302), the second limiting wheel (303), and the supporting wheel (304) are distributed in a triangular shape, and the first limiting wheel (302) and the second limiting wheel (303) are arranged on the same side of the optical cable.

7. The optical cable sheath defect recognition device based on machine vision according to claim 6, characterized in that: It also includes a guide mechanism (400), which includes support plates (401) arranged on both sides of the shell (100), and a first guide wheel (402) arranged on the support plate (401), and a second guide wheel (403) is arranged on the support plate (401); The first guide wheel (402) and the second guide wheel (403) are respectively arranged on both sides of the optical cable.

8. A machine vision-based optical cable sheath defect recognition system, comprising the machine vision-based optical cable sheath defect recognition device according to any one of claims 1 to 7, characterized in that: Also includes, A position recording module, which includes a speed encoder and a meter recorder; A communication module, which is used for data transmission; The sound and light alarm module is used to send out different sound and light signals according to the detection results.

9. The optical cable sheath defect recognition system based on machine vision as claimed in claim 8, characterized in that: The speed encoder and meter recorder are used to measure the moving speed and length data of the optical cable.

10. A method for machine vision training of optical cable defects, applied to the optical cable sheath defect recognition system based on machine vision according to claim 8 or 9, characterized in that: include, Collect different types of optical cable defect images; Build defect target detection model; Training the recognition model; Build the trained model.

Images