Cable threading equipment for cable processing

Abstract

The invention discloses cable threading equipment for cable processing, which comprises an extruder, a cooler and a traction device which are connected in sequence, and also comprises a detection device which is arranged between the cooler and the traction device and comprises an image processing mechanism and a plurality of image acquisition mechanisms, the plurality of image acquisition mechanisms are uniformly arranged along the circumference of the cable, each image acquisition mechanism is electrically connected with the image processing mechanism, the image acquisition mechanisms are used for acquiring a surface image of the cable, and the image processing mechanism is used for extracting features of the surface image and fusing the features of the surface image of each image acquisition mechanism to form a fused image; and extracting and detecting features of the fused image. According to the method, the image of the cable is collected, then the insulating layer of the cable is detected according to the image, target nondestructive detection is achieved, manpower is not consumed in the process, and the operation efficiency is high.

Description

Technical Field

[0001] The invention relates to the technical field of cable threading, and in particular to a cable threading device for cable processing. Background Art

[0002] During the production process of the cable, an insulating layer is formed on the surface of the cable after passing through the extruder, and then cooled by the cooler to make the plastic insulating layer solidify quickly so that the cable can be quickly stored. Therefore, the detection of defects in the insulation layer of the cable is very important. At present, the detection of local insulation defects of the cable is generally carried out through manual observation during the cable production process, or after the production and storage of the cable is completed, the cable is tested by discharge detection. The above two methods are labor-intensive and lack intelligence, which seriously affects the detection of the insulation layer of the cable and the production efficiency. Summary of the invention

[0003] The purpose of the present invention is to provide a cable threading equipment for cable processing to address the above-mentioned problems. The equipment can capture images of the cables and then detect the insulation layer of the cables based on the images to achieve target non-destructive testing. This process does not consume manpower and has high operating efficiency.

[0004] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is as follows:

[0005] According to one aspect of the present invention, there is provided a cable threading device for cable processing, comprising an extruder, a cooler and a traction device, wherein the extruder, the cooler and the traction device are connected in sequence;

[0006] Also included is a detection device, the detection device is arranged between the cooler and the traction device;

[0007] The detection device includes an image processing mechanism and a plurality of image acquisition mechanisms;

[0008] A plurality of the image acquisition mechanisms are evenly arranged along the circumference of the cable;

[0009] Each of the image acquisition mechanisms is electrically connected to the image processing mechanism;

[0010] The image acquisition mechanism is used to acquire the surface image of the cable; the image processing mechanism is used to extract the features of the surface image, fuse the features of the surface images of each image acquisition mechanism to form a fused image, and extract and detect the features of the fused image.

[0011] Preferably, the image processing mechanism includes a microcontroller and a storage medium, and the microcontroller is electrically connected to the storage medium.

[0012] Preferably, the microcontroller includes a main control unit and an image processing unit, the main control unit is electrically connected to the image processing unit, the main control unit is used to retrieve image data from a storage medium and transmit the image data to the image processing unit, and the image processing unit is used to process the image data.

[0013] Preferably, the image processing unit includes an image preprocessing module, an edge detection module, a feature extraction module and a feature fusion module;

[0014] The image preprocessing module is used to preprocess the image; the edge detection module is used to detect the edge information of the image; the feature extraction module is used to extract the features of the image; and the feature fusion module is used to fuse the extracted features to obtain a fused image.

[0015] Preferably, the image preprocessing module includes a denoising submodule and a contrast submodule, the denoising submodule is used to reduce noise in the image, and the contrast submodule is used to adjust the contrast of the image.

[0016] Preferably, the feature extraction module includes a model building submodule, a training set generation submodule and a model training submodule;

[0017] The model construction submodule is used to construct a feature extraction model; the training set generation submodule is used to generate a training set for training the feature extraction model; and the model training submodule is used to train the feature extraction model using the training set.

[0018] Preferably, the image acquisition mechanism includes a CCD camera and an analog-to-digital converter, the CCD camera is electrically connected to the analog-to-digital converter, and the analog-to-digital converter is electrically connected to the microcontroller.

[0019] Preferably, the traction device comprises a rotating motor and a drum for receiving the cable, and the output end of the rotating motor is fixedly connected to the drum.

[0020] Preferably, it also includes a marking device, which includes a laser generator, a laser controller and an encoder, the laser generator is arranged between the detection device and the traction device, the laser generator is electrically connected to the laser controller, the encoder is connected to the rotating motor, and the encoder is electrically connected to the laser controller.

[0021] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0022] The present invention arranges a detection device between the cooler and the traction device. The detection device includes an image processing mechanism and a plurality of image acquisition mechanisms. The image acquisition mechanism is used to collect images of the cable, and then the image processing mechanism is used to detect whether there are defects in the image of the insulation layer of the cable, thereby achieving target non-destructive detection. This process does not consume manpower and has high operating efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a schematic diagram of the structure of the present invention;

[0024] Figure 2 is a structural diagram of an image processing mechanism of the present invention;

[0025] Figure 3 is a structural diagram of an image processing unit of the present invention;

[0026] Figure 4 It is a structural diagram of the image acquisition mechanism of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and preferred embodiments. However, it should be noted that many details listed in the specification are only for the purpose of enabling the reader to have a thorough understanding of one or more aspects of the invention, and these aspects of the present invention can be realized even without these specific details.

[0028] See also Figures 1 to 4 The present invention provides a cable threading device for cable processing, and the technical solution is as follows:

[0029] A cable threading device for cable processing includes an extruder, a cooler and a traction device, which are connected in sequence. The extruder is used to attach an insulating layer to the outer periphery of the cable. When the cable passes through the extruder, its outer surface is covered by the plastic melted by the extruder to form an insulating layer that wraps the cable. The cable attached with the insulating layer through the extruder needs to be cooled by the cooler to make the plastic solidify quickly. The cooler generally uses circulating water to cool the cable. The cooled cable is received by the traction device, and the traction device is also used to drive the cable to move. The traction device includes a rotating motor and a roller for receiving the cable. The output end of the rotating motor is fixedly connected to the roller, and the cable is wound around the roller. The roller is driven to rotate by the rotating motor. The cooled cable is received under the rotation of the roller and is continuously moved to ensure the operation of the whole set of equipment.

[0030] In order to prevent the insulation layer of the cable from being damaged, there is a risk of leakage short circuit. Therefore, in this embodiment, a detection device is set between the cooler and the traction device to detect the surface of the cable. Specifically, the detection device includes an image processing mechanism and a plurality of image acquisition mechanisms. The image processing mechanism includes a microcontroller and a storage medium, and the microcontroller is electrically connected to the storage medium. The storage medium can be a mechanical hard disk, a solid-state hard disk, a mobile USB flash drive, etc., and the devices that can be used to store data are all storage media described in this application, and the storage medium is used to store the image data of the cable. The microcontroller includes a main control unit and an image processing unit. The main control unit includes a main control chip, which is a central processing unit. The main control chip can retrieve image data from the storage medium, and then transmit the image data to the image processing unit, and use the image processing unit to process the image data. The image processing unit is a GPU chip, including an image preprocessing module, an edge detection module, a feature extraction module and a feature fusion module.

[0031] The image preprocessing module is used to preprocess the image. The image preprocessing module includes a denoising submodule and a contrast submodule. The denoising submodule is used to reduce the noise in the image, and the contrast submodule is used to adjust the contrast of the image. The image preprocessing specifically includes the following steps:

[0032] Image reading: reading images from storage media for subsequent processing;

[0033] Filtering and denoising: In the denoising submodule, filters (such as mean filter, median filter, Gaussian filter, etc.) are used to remove noise from the image to prevent noise from affecting the image quality and the effect of subsequent processing;

[0034] Contrast enhancement: In the contrast submodule, the contrast of the image is adjusted by histogram equalization, contrast stretching and other methods to make the useful information in the image more prominent. The purpose of contrast enhancement is to highlight the points in the image where the grayscale has changed significantly (i.e., potential edge points). This can be done by calculating the gradient amplitude. The gradient is a vector that represents the direction and rate of brightness change in the image. The gradient amplitude reflects the intensity of the brightness change, so the edge information can be enhanced by calculating the gradient amplitude of each pixel.

[0035] By preprocessing the image, irrelevant information in the image is eliminated, useful real information is restored, and the detectability of relevant information, such as edges, textures and other features, is enhanced. The data is simplified to the maximum extent, the amount of calculation for subsequent processing is reduced, and the reliability of feature extraction, image segmentation, matching and recognition is improved.

[0036] Because the edge is the place where the grayscale changes most dramatically in the image, it usually corresponds to the outline of the object. Therefore, in order to extract useful information from the image, the edge information is extracted from the preprocessed image by the edge detection module. Specifically, the commonly used edge detection algorithm detects edge points. The edge detection algorithm includes methods such as the Sobel operator and the Canny operator. In this embodiment, the edge points are detected by methods such as thresholding. That is, one or more thresholds are set, and points with gradient amplitudes greater than the thresholds are regarded as edge points. After the edge points are detected, the sub-pixel edge positioning technology is used to further determine the exact position of the edge. The sub-pixel edge positioning technology can more accurately locate the edge position below the pixel level, thereby improving the accuracy of edge detection.

[0037] After the image is detected by the edge detection module, the feature extraction module performs feature extraction. Specifically, the feature extraction module includes a model construction submodule, a training set generation submodule and a model training submodule. The model construction submodule is used to build a suitable extraction model. In this embodiment, a convolutional neural network CNN model is used as an image feature extraction model. The CNN model extracts hierarchical features in the image through structures such as convolutional layers, pooling layers, and fully connected layers. The convolutional layer is responsible for learning local features (such as edges and textures), the pooling layer is responsible for reducing the size of the feature map and learning global features (such as shape and size), and the fully connected layer is responsible for converting the learned features into classification or recognition results.

[0038] The steps of feature extraction are as follows:

[0039] 1. Data Preparation

[0040] Data collection: First, a large amount of image data needs to be collected as training sets and test sets, ensuring that the data is representative and can cover various situations of the target task.

[0041] Data preprocessing: First, ensure that all image data is converted to a unified format; then normalize the pixel values ​​of the image to a specific range of [0,1] or [-1,1] to speed up training and improve model performance; then increase the diversity of the data through random cropping, rotation, flipping and other operations to improve the generalization ability of the model; finally, divide the data set into training set, validation set and test set. Usually, the training set is used to train the model, the validation set is used to adjust hyperparameters and monitor overfitting, and the test set is used to evaluate the final performance of the model.

[0042] 2. Model Construction

[0043] Select the CNN network structure, which includes the number and parameter settings of the input layer, convolution layer, pooling layer, and fully connected layer.

[0044] Input layer: Receives image data and represents the image data as a multi-channel two-dimensional matrix, where each channel corresponds to a feature of the image, such as red, green, and blue channels representing a color image.

[0045] Convolutional layer: Determine the number, size, step size, and padding of convolution kernels for each convolutional layer. During the convolution operation, a set of learnable convolution kernels (also called filters) are slid over the input image and convoluted with the local area of ​​the image to generate a feature map. Each convolution kernel is responsible for extracting a specific feature in the image, such as edges, textures, etc. After the convolution operation, a nonlinear activation function (such as ReLU) is usually used to activate the feature map to increase the nonlinear expression ability of the network. And by stacking multiple convolutional layers, higher-level features in the image are extracted layer by layer.

[0046] Pooling layer: Determine the type of pooling layer and the size and step size of the pooling window. After the convolution layer, the pooling layer is used to reduce the spatial size of the feature map, reduce the amount of calculation, and improve the model's ability to resist noise. Commonly used pooling operations include maximum pooling and average pooling, which retain key feature information by selecting the maximum value or average value in a local area while ignoring unimportant details.

[0047] After the convolutional and pooling layers, the feature maps are flattened into one-dimensional vectors for input into the fully connected layers.

[0048] Fully connected layer: Each neuron in the fully connected layer is connected to all neurons in the previous layer, linearly combined through weights and biases, and nonlinearly activated using activation functions. The fully connected layer is responsible for mapping the extracted features to the final output category.

[0049] Classification output: The final fully connected layer outputs the predicted probability or score of each category, which is normalized by activation functions such as the softmax function to obtain the final classification result.

[0050] 3. Model Training

[0051] Initialize the weights and biases in the network using random values ​​or specific initialization methods (such as He initialization, Xavier initialization, etc.). Input the training set data into the network, and finally get the prediction results through layer-by-layer calculations of the convolution layer, activation function layer, and pooling layer. Calculate the loss value based on the prediction results and the true label to evaluate the prediction performance of the model. Calculate the gradient of the loss function to the network parameters through the backpropagation algorithm. Use the optimizer to update the network parameters according to the gradient to reduce the loss value. Repeat the process of forward propagation, loss calculation, backpropagation, and parameter update until the preset number of iterations is reached or other stopping conditions are met (such as the performance of the validation set no longer improves).

[0052] IV. Model Evaluation and Optimization

[0053] Model evaluation: Use the test set to evaluate the final performance of the model, such as accuracy, recall, F1 score and other indicators. The trained CNN model can be used for image feature extraction.

[0054] After image feature extraction, the feature fusion module is used to fuse the features. Specifically, feature map fusion: After extracting feature maps of different images or different parts of the same image, it is necessary to use a specific fusion strategy to fuse these feature maps. Common fusion strategies include element-wise sum, tensor concatenation, etc. Element-wise sum: Add the feature values ​​of corresponding positions, which is applicable to the case where feature maps have the same size and number of channels. Tensor concatenation: Concatenate different feature maps in the channel dimension, which is applicable to the case where feature maps have different numbers of channels but the same size. Attention mechanism (optional): In order to further improve the fusion effect, an attention mechanism can be introduced in the fusion process. The attention mechanism can dynamically adjust the fusion weights according to the importance of different features, so that the model pays more attention to important feature information. IV. Feature reconstruction and post-processing Feature reconstruction: After feature fusion is completed, it may be necessary to reconstruct the fused feature map into an image of the same size as the original image through structures such as deconvolution layers and upsampling layers. Post-processing: Finally, the reconstructed image can be post-processed (such as smoothing, sharpening, etc.) to further improve the image quality and visual effect.

[0055] After the feature fusion is completed, a fused image is obtained, and the fused image is tested to detect whether there are defects on the fused image.

[0056] Among them, the image of the cable is obtained by an image acquisition mechanism. Specifically, the image acquisition mechanism includes a CCD camera and an analog-to-digital converter. The CCD camera is used to capture the surface image of the cable. The CCD camera is electrically connected to the analog-to-digital converter. The CCD camera obtains the surface image of the cable and transmits the surface image to the analog-to-digital converter. The analog-to-digital converter converts the surface image into an image digital signal. The analog-to-digital converter is electrically connected to a microcontroller. The analog-to-digital conversion transmits the image digital signal to the microcontroller, and the microcontroller is used to identify and process the image digital signal.

[0057] In order to realize the automation of the cable production line, the cable is received in the traction device, and at the same time, the cable can be moved in the production line through the traction device. Specifically, the traction device includes a rotating motor and a drum for receiving the cable, and the output end of the rotating motor is fixedly connected to the drum. The rotating motor drives the drum to rotate, and the prepared cable is wound on the drum, and the cable is received by the drum, and at the same time, a driving force is provided to the cable.

[0058] Furthermore, in order to mark the defective location in time after a defect occurs during cable detection, in this embodiment, a marking device is provided between the detection device and the traction device. The marking device includes a laser generator, a laser controller and an encoder. The laser generator is electrically connected to the laser controller, and the laser controller is used to control the operation of the laser generator. The laser generator is used to mark the defective location on the surface of the insulation layer of the cable so that the defective location can be quickly found later. The encoder is connected to the rotating motor, and the encoder is electrically connected to the laser controller. The encoder reads the rotation distance of the rotating motor to obtain the distance the cable moves. In conjunction with the above-mentioned cable defect detection, by detecting the length of the cable defect, the laser generator marks the surface of the cable twice according to the length of the defect. Through the two markings, the defect is within the range of the two markings, so that the defect can be found later, thereby improving the production efficiency of the cable.

[0059] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A cable threading device for cable processing, characterized in that: It comprises an extruder, a cooler and a traction device, wherein the extruder, the cooler and the traction device are connected in sequence; Also included is a detection device, the detection device is arranged between the cooler and the traction device; The detection device includes an image processing mechanism and a plurality of image acquisition mechanisms; A plurality of the image acquisition mechanisms are evenly arranged along the circumference of the cable; Each of the image acquisition mechanisms is electrically connected to the image processing mechanism; The image acquisition mechanism is used to acquire the surface image of the cable; the image processing mechanism is used to extract the features of the surface image, fuse the features of the surface images of each image acquisition mechanism to form a fused image, and extract and detect the features of the fused image.

2. The cable threading device for cable processing according to claim 1, characterized in that: The image processing mechanism includes a microcontroller and a storage medium, and the microcontroller is electrically connected to the storage medium.

3. The cable threading device for cable processing according to claim 2, characterized in that: The microcontroller includes a main control unit and an image processing unit. The main control unit is electrically connected to the image processing unit. The main control unit is used to retrieve image data from a storage medium and transmit the image data to the image processing unit. The image processing unit is used to process the image data.

4. The cable threading device for cable processing according to claim 3, characterized in that: The image processing unit includes an image preprocessing module, an edge detection module, a feature extraction module and a feature fusion module; The image preprocessing module is used to preprocess the image; the edge detection module is used to detect the edge information of the image; the feature extraction module is used to extract the features of the image; and the feature fusion module is used to fuse the extracted features to obtain a fused image.

5. The cable threading device for cable processing according to claim 4, characterized in that: The image preprocessing module includes a denoising submodule and a contrast submodule. The denoising submodule is used to reduce noise in the image, and the contrast submodule is used to adjust the contrast of the image.

6. The cable threading device for cable processing according to claim 4, characterized in that: The feature extraction module includes a model building submodule, a training set generation submodule and a model training submodule; The model building submodule is used to build a feature extraction model; The training set generation submodule is used to generate a training set for training a feature extraction model; and the model training submodule is used to train the feature extraction model using the training set.

7. The cable threading device for cable processing according to claim 2, characterized in that: The image acquisition mechanism comprises a CCD camera and an analog-to-digital converter, wherein the CCD camera is electrically connected to the analog-to-digital converter, and the analog-to-digital converter is electrically connected to the microcontroller.

8. The cable threading device for cable processing according to claim 1, characterized in that: The traction device comprises a rotating motor and a drum for receiving the cable, and the output end of the rotating motor is fixedly connected to the drum.

9. The cable threading device for cable processing according to claim 8, characterized in that: It also includes a marking device, which includes a laser generator, a laser controller and an encoder. The laser generator is arranged between the detection device and the traction device, the laser generator is electrically connected to the laser controller, the encoder is connected to the rotating motor, and the encoder is electrically connected to the laser controller.

Images