Method and system for detecting clamping state abnormity of cable laid along railway

Abstract

The invention provides a clamping state anomaly detection method and system for a cable laid along a railway. According to the anomaly detection method, a training data set is constructed, a cable anomaly detection model is pre-trained, a cable anomaly detection pre-training model can adapt to various universal environments, data generated by a to-be-detected cable is combined to form an experience pool, and the pre-trained cable anomaly detection model is finely adjusted. According to the method, the cable anomaly detection model is updated, the method is more suitable for the environment where the to-be-detected cable is located, the accuracy of the anomaly detection method of the cable clamp is improved, in addition, the method can be further suitable for anomaly detection of the to-be-detected cable in combination with fine adjustment of the confidence coefficient, and the accuracy of the anomaly detection method of the cable clamp is further improved. According to the invention, the accuracy and real-time performance of fault detection of the leaky coaxial cable can be effectively improved, intelligent state monitoring and prediction maintenance are realized, and the method has a wide application prospect.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence, and more particularly to the field of automatic cable detection, specifically to a method and system for detecting abnormal clamping states of cables laid along railway lines. Background Technology

[0002] In cable laying projects, some cables require long-distance overhead installation according to requirements, secured with cable clamps at intervals. In railway cable laying requirements, leaky coaxial cables, electrical cables, and fiber optic cables need to be laid long distances along the track. During installation, according to construction specifications, cable clamps are used at intervals (generally 1m) to suspend the cables on supports or fix them to the tunnel walls. These cable clamps are exposed to harsh environments such as moisture and vibration for extended periods, making them prone to aging, loosening, or breakage. Cable detachment from the clamps poses a significant safety risk and requires inspection to prevent potential safety hazards to trains and other systems.

[0003] Current technologies for monitoring whether cables have detached from cable clamps primarily rely on manual inspections. For example, periodic inspections of GSM-R leaky cables are labor-intensive, inefficient, and involve prolonged handheld lighting and upward-looking inspections, posing safety hazards. Furthermore, manual inspections are time-consuming, have limited coverage, and are prone to overlooking potential issues. In contrast, a more intelligent approach involves deploying image recognition devices on each clamp. However, this increases equipment costs and requires a vehicle to pass over the clamp location to collect images and determine if a detachment has occurred.

[0004] Therefore, monitoring whether the cable has detached from the clamp requires consideration of many factors such as cost, effectiveness, and efficiency. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for detecting abnormal clamping states of cables laid along railway lines, which mainly solves the problem of inaccurate detection of clamp detachment from cables.

[0006] The objective of this invention is achieved through the following methods:

[0007] According to a first aspect of the present invention, a method for detecting abnormal clamping states of cables laid along railway lines is provided. The method for detecting abnormalities in cable clamps includes: S1, constructing a training dataset and pre-training a cable abnormality detection model using the training dataset to obtain a pre-trained cable abnormality detection model; wherein the training dataset includes multiple samples, each sample consisting of a cable vibration signal, a label corresponding to the cable vibration signal, and the location of clamp detachment; the cable vibration signal is the vibration signal generated by the cable during a predetermined length of laying and after a fixed number of clamps are used; the cable abnormality detection model includes at least a convolutional neural network and a classifier; S2, acquiring the vibration signal generated by the cable to be detected during a predetermined length of laying, and performing preliminary abnormality detection of the cable clamps using the pre-trained cable abnormality detection model obtained in S1; when the cable clamps of the cable to be detected show an abnormality, the cable clamps are... S3. A new sample is formed by combining the vibration signal and clamp anomaly label corresponding to the cable under test when an anomaly is initially detected. This new sample is then combined with the training dataset to form an experience pool. S4. The proportion of the new sample in the experience pool in S2 is determined, and the confidence level in the cable anomaly detection pre-training model is adjusted using this proportion. The cable anomaly detection pre-training model after confidence level adjustment is then fine-tuned using the experience pool to obtain the final cable anomaly detection model. S5. The final cable anomaly detection model is used to detect anomalies in the vibration signal generated by the cable under test during the laying of a preset length. The clamp detachment position corresponding to the cable under test when an anomaly is detected is determined by the similarity between the vibration signal corresponding to the cable under test when an anomaly is detected and each sample in the training dataset.

[0008] In some embodiments of the present invention, in the anomaly detection method, the location of the clamp detachment in the training dataset is the location where no clamp is set in the cable during the laying of a preset length, and the label corresponding to the cable vibration signal includes normal fixing label and abnormal fixing label.

[0009] In some embodiments of the present invention, in the anomaly detection method, when pre-training or fine-tuning the cable anomaly detection model, the parameters of the cable anomaly detection model are updated using the following cross-entropy loss:

[0010]

[0011] in, This indicates an abnormal fixed label. Represents cross-entropy loss, This indicates the probability that the cable vibration signal is assigned to an abnormal fixed label. When fine-tuning the cable anomaly detection model, the learning rate of the cable anomaly detection model is set to 0.001.

[0012] In some embodiments of the present invention, in S3, the confidence level in the pre-trained model for cable anomaly detection is adjusted as follows: based on the increase in the proportion of new samples in the experience pool, the confidence level corresponding to the probability that the cable vibration signal is classified into an abnormal fixed label is increased.

[0013] Furthermore, in the anomaly detection method, the confidence level corresponding to the probability of the cable vibration signal being assigned to an abnormal fixed label is improved in the following ways: when the proportion of new samples in the experience pool is greater than 5% and less than 10%, the confidence level corresponding to the probability of the cable vibration signal being assigned to an abnormal fixed label is 90%; when the proportion of new samples in the experience pool is greater than or equal to 10%, the confidence level corresponding to the probability of the cable vibration signal being assigned to an abnormal fixed label is 95%.

[0014] In some embodiments of the present invention, in step S4, the similarity between the vibration signal of the cable to be detected when an anomaly is detected and each sample in the training dataset is determined in the following manner:

[0015] The similarity between the vibration signal of the cable under test when an anomaly is detected and the vibration signal of each sample in the training dataset is calculated.

[0016] The label corresponding to the largest similarity among all calculated similarities is used as the corresponding clamp detachment location of the cable under test when an anomaly is detected.

[0017] In some embodiments of the present invention, in step S4, the similarity between the vibration signal of the cable to be detected when an anomaly is detected and each sample in the training dataset is determined in the following manner:

[0018] The similarity between the vibration signal of the cable under test when an anomaly is detected and the vibration signal of each sample in the training dataset is calculated. All the calculated similarities are sorted from largest to smallest, and the clamp detachment position labels corresponding to the top N positions in the sort are taken as the clamp detachment position of the cable under test when an anomaly is detected. The number of times the cable anomaly detection model is fine-tuned is determined.

[0019] According to a second aspect of the present invention, an anomaly detection system for cable clamps is provided, the anomaly detection system comprising:

[0020] A vibration sensor is configured to collect vibration signals generated after a cable is laid in a designated area.

[0021] A server, connected to the vibration sensor, is configured to perform the steps of the method as described in the first aspect of the invention.

[0022] Compared with existing technologies, the present invention has the following beneficial effects: By constructing a training dataset, the present invention pre-trains the cable anomaly detection model, enabling the pre-trained model to adapt to various common environments. Furthermore, by combining the data generated by the cables to be detected into an experience pool, the pre-trained cable anomaly detection model is fine-tuned, allowing for updates and better adaptation to the environment of the cables to be detected, thus improving the accuracy of the cable clamp anomaly detection method. Moreover, by combining confidence level fine-tuning, it can further adapt to the anomaly detection of the cables to be detected, further improving the accuracy of the cable clamp anomaly detection method. Attached Figure Description

[0023] The embodiments of the present invention will be further described below with reference to the accompanying drawings:

[0024] Figure 1 This is a flowchart illustrating a method for detecting abnormal clamping states of cables laid along railway lines, as provided in an embodiment of the present invention.

[0025] Figure 2 This is a schematic diagram of the structure of a cable anomaly detection model provided in an embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and are not intended to limit the invention.

[0027] As mentioned in the background section, "existing monitoring of cable clamps relies on manual inspection methods. For example, periodic inspections of GSM-R leaky cables are not only labor-intensive and inefficient, but also pose safety hazards due to prolonged handheld lighting and upward-looking inspections. Furthermore, manual inspections are time-consuming and have limited coverage, easily overlooking localized hazards. Additionally, the deployment of image recognition intelligent devices on each clamp increases equipment costs for identifying anomalies." Therefore, this invention delves into an anomaly detection method for cable clamps, performing anomaly detection with reduced sensor setup. It fully utilizes artificial intelligence's ability to perceive the information contained in the vibration signals collected by sensors, improving the accuracy and robustness of the anomaly detection method. In this embodiment, the clamp detachment location includes situations such as the cable completely detaching from the clamp, the clamp detaching, or the clamp not exerting any binding force on the cable.

[0028] Specifically, the present invention is as follows Figure 1As shown, a method for detecting abnormal clamping states of cables laid along railway lines is provided, mainly including: constructing a training dataset and pre-training a cable anomaly detection model; forming an experience pool; setting confidence in the pre-trained model and fine-tuning the cable anomaly detection model; and anomaly detection. According to an embodiment of the present invention, a method for detecting abnormal clamping states of cables laid along railway lines is provided. The cable clamp anomaly detection method includes: S1, constructing a training dataset and pre-training a cable anomaly detection model using the training dataset to obtain a cable anomaly detection pre-trained model; wherein, the training dataset includes multiple samples, each sample consisting of a cable vibration signal, a label corresponding to the cable vibration signal, and the clamp detachment position; the cable vibration signal is the vibration signal generated by the cable during a preset length of laying and after a fixed number of clamps; the cable anomaly detection model includes at least a convolutional neural network and a classifier; S2, acquiring the vibration signal generated by the cable to be detected during a preset length of laying, and using the cable anomaly detection pre-trained model obtained in S1 to perform preliminary anomaly detection of the cable clamps; when the cable clamps of the cable to be detected show an anomaly, the... S3. A new sample is formed by combining the vibration signal and clamp anomaly label corresponding to the cable under test when an anomaly is initially detected. This new sample is then combined with the training dataset to form an experience pool. S4. The proportion of the new sample in the experience pool in S2 is determined, and the confidence level in the cable anomaly detection pre-training model is adjusted using this proportion. The cable anomaly detection pre-training model after confidence level adjustment is then fine-tuned using the experience pool to obtain the final cable anomaly detection model. S5. The final cable anomaly detection model is used to detect anomalies in the vibration signal generated by the cable under test during the laying of a preset length. The clamp detachment position corresponding to the cable under test when an anomaly is detected is determined by the similarity between the vibration signal corresponding to the cable under test when an anomaly is detected and each sample in the training dataset.

[0029] In this embodiment of the invention, a pre-trained cable anomaly detection model is obtained through pre-training the cable anomaly detection model. Then, incremental processing of the training data using this pre-trained model and the cable to be detected is performed. This improves detection accuracy while adapting to the specific environment of the cable to be detected. The experience pool obtained from the incremental processing of the training data is used to fine-tune the confidence-adjusted pre-trained cable anomaly detection model, resulting in the final cable anomaly detection model. This further improves the overall accuracy of the detection method.

[0030] To better illustrate this invention, let's first explain the cable anomaly detection model used in this invention. The cable anomaly detection model can use convolutional neural networks (CNNs) and their variants combined with SVMs, random forests (RFs), or XGBoost classifiers to improve classification accuracy; variants of convolutional neural networks include one-dimensional convolutional neural networks (1D-CNNs), residual neural networks (DRISNs), and improved deep convolutional neural networks (DCNNs), etc.

[0031] For ease of explanation, the cable anomaly detection model used in this invention is as follows: Figure 2 The diagram illustrates the detection method using a CNN + fully connected layer (classifier) ​​structure. It's important to note that the cable vibration signal input to the cable anomaly detection model can be preprocessed using multi-scale wavelet decomposition (WPD) to effectively separate the signal and remove noise; then, short-time Fourier transform (STFT) is used to extract time-frequency features as input, improving classification accuracy. This process primarily reduces interference from moving trains on the track, thereby enhancing the accuracy of the detection method.

[0032] The following describes in detail each step of a method for detecting abnormal clamping conditions of cables laid along railway lines.

[0033] In step S1, a training dataset is constructed, and the cable anomaly detection model is pre-trained using the training dataset to obtain a pre-trained cable anomaly detection model. The training dataset includes multiple samples, each sample consisting of a cable vibration signal, a label corresponding to the cable vibration signal, and the location of the clamp detachment. The cable vibration signal is the vibration signal generated when the cable is laid for a preset length and the number of clamps is fixed. The cable anomaly detection model consists of at least a convolutional neural network and a classifier.

[0034] Preferably, each sample in the training dataset is generated after the cable has been laid to a predetermined length, by detecting the cable vibration signal after it detaches from a clamp. This signal is considered as data for one sample. Alternatively, the data for each sample can be understood as including the cable vibration signal, the location of the unique clamp detachment, and any abnormally fixed tags. This facilitates locating the clamp detachment point. Of course, the cable vibration signal can be actively triggered by the same vibration source. The preferred laying length is 5-10 kilometers.

[0035] In step S2, vibration signals generated by the cable under test during its laying along a preset length are acquired, and the cable clamps are pre-trained using the cable anomaly detection model obtained in step S1 for preliminary anomaly detection. When the cable clamps of the cable under test show an anomaly during the preliminary detection, the vibration signal corresponding to the cable under test at the time of the anomaly and the clamp anomaly label are combined to form a new sample, and this new sample is combined with the training dataset to form an experience pool. This preliminary anomaly detection expands the training dataset, improving the granularity of the training data and enhancing the accuracy of the overall detection method while adapting to the usage environment of the cable under test. According to an embodiment of the present invention, in the anomaly detection method, the clamp detachment position in the training dataset is the position where no clamps are installed during the laying of the cable along the preset length, and the labels corresponding to the cable vibration signals include normal fixing labels and abnormal fixing labels.

[0036] In step S3, the proportion of new samples in the experience pool in step S2 is determined, and the confidence level in the pre-trained model for cable anomaly detection is adjusted using this proportion; and the cable anomaly detection pre-trained model with adjusted confidence level is fine-tuned using the experience pool to obtain the final cable anomaly detection model.

[0037] According to one embodiment of the present invention, in S3, the confidence level in the pre-trained model for cable anomaly detection is adjusted as follows: based on the increase in the proportion of new samples in the experience pool, the confidence level corresponding to the probability that the cable vibration signal is classified under the abnormal fixed label is increased. Further, in the anomaly detection method, the confidence level corresponding to the probability that the cable vibration signal is classified under the abnormal fixed label is increased as follows: when the proportion of new samples in the experience pool is greater than 5% and less than 10%, the confidence level corresponding to the probability that the cable vibration signal is classified under the abnormal fixed label is 90%; when the proportion of new samples in the experience pool is greater than or equal to 10%, the confidence level corresponding to the probability that the cable vibration signal is classified under the abnormal fixed label is 95%.

[0038] It's important to note that each sample in the entire experience pool is discretely distributed and does not affect the model's training and convergence. Therefore, when fine-tuning the pre-trained cable anomaly detection model with confidence adjustments using the experience pool, batch training with replacement can be employed to reduce sampling difficulty during training. As the proportion of new samples in the experience pool increases, the learning rate can be gradually increased and decreased to adapt to the hyperparameters of the cables to be detected, such as confidence, thereby increasing the model's accuracy.

[0039] In the above steps, according to an embodiment of the present invention, in the anomaly detection method, when pre-training or fine-tuning the cable anomaly detection model, the parameters of the cable anomaly detection model are updated using the following cross-entropy loss:

[0040]

[0041] in, This indicates an abnormal fixed label. Represents cross-entropy loss, This indicates the probability that the cable vibration signal is assigned to an abnormal fixed label. When fine-tuning the cable anomaly detection model, the learning rate of the cable anomaly detection model is set to 0.001.

[0042] In step S4, the final cable anomaly detection model is used to detect the vibration signal generated by the cable under test during the laying of a preset length. The location of the clamp falling off when the cable under test detects an anomaly is determined by the similarity between the vibration signal of the cable under test when the anomaly is detected and each sample in the training dataset.

[0043] According to one embodiment of the present invention, in S4, the similarity between the vibration signal of the cable to be detected when an anomaly is detected and each sample in the training dataset is determined in the following manner:

[0044] The similarity between the vibration signal of the cable under test when an anomaly is detected and the vibration signal of each sample in the training dataset is calculated. The label of the clamp detachment location corresponding to the highest similarity among all calculated results is taken as the clamp detachment location of the cable under test when an anomaly is detected. It should be noted that when the cable anomaly detection model has a high number of fine-tuning times, the detection accuracy of the model is improved. Using the clamp detachment location label corresponding to the highest similarity can be used as the clamp detachment location of the cable under test when an anomaly is detected, which is convenient for maintenance.

[0045] In some embodiments of the present invention, in step S4, the similarity between the vibration signal of the cable to be detected when an anomaly is detected and each sample in the training dataset is determined in the following manner:

[0046] The similarity between the vibration signal of the cable under test when an anomaly is detected and the vibration signal of each sample in the training dataset is calculated. All similarities are then sorted from largest to smallest. The top N similarities correspond to the clamp detachment location labels as the clamp detachment locations of the cable under test when an anomaly is detected. N is determined based on the number of fine-tuning iterations of the cable anomaly detection model. Specifically, the value of N is inversely proportional to the number of fine-tuning iterations; for example, N is 6 for 6 iterations and 5 for 7 iterations. This ensures that with fewer fine-tuning iterations, determining multiple clamp detachment locations reduces the clamp detection area to some extent and improves the accuracy of the detection method.

[0047] According to a second aspect of the present invention, an anomaly detection system for cable clamps is provided, the anomaly detection system comprising:

[0048] The system includes a vibration sensor and a server. The vibration sensor is configured to collect vibration signals generated by a cable laid in a designated area. The server, connected to the vibration sensor, is configured to perform the steps of the method described in the first aspect of the invention. The server and the vibration sensor can be connected via a wireless unit such as 5G. The server is primarily used for model deployment and inference calculations. The anomaly detection system filters and transforms the vibration signals generated by the cable under test during its layup of a preset length, and inputs these signals into the cable anomaly detection model for training and evaluation. When the model detects an anomaly or predicts a potential fault, the system will issue a warning through an alarm device or monitoring platform, prompting maintenance personnel to check and handle the issue promptly.

[0049] This invention provides a method and system for detecting abnormal clamping status of cables laid along railway lines, which can effectively improve the accuracy and real-time performance of leaky cable fault detection, realize intelligent status monitoring and predictive maintenance, and has broad application prospects.

[0050] This invention can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of the invention.

[0051] Computer-readable storage media can be tangible devices that hold and store instructions for use by an instruction execution device. Computer-readable storage media can be, for example, including but not limited to, electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination thereof.

[0052] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or technical improvements to the embodiments in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method of detecting abnormality in a clamping state of a railway line cable laying clamp, characterized by, The abnormality detection method comprises: S1, constructing a training data set, and pre-training a cable abnormality detection model by using the training data set to obtain a cable abnormality detection pre-training model; wherein the training data set comprises a plurality of samples, each sample is composed of a cable vibration signal, a label corresponding to the cable vibration signal and a clamp shedding position, the cable vibration signal is a vibration signal generated by a cable in a preset length laying and fixed by a set number of clamps; the cable abnormality detection model at least comprises a convolutional neural network and a classifier; S2, obtaining a vibration signal generated by a to-be-detected cable in a preset length laying, and performing preliminary cable clamp abnormality detection by using the cable abnormality detection pre-training model obtained in S1; When the to-be-detected cable clamp preliminary detection is abnormal, a new sample is composed of the vibration signal corresponding to the to-be-detected cable in the preliminary detection abnormality and the clamp abnormality label, and the new sample and the training data set form an experience pool; S3, determining the proportion of the new sample in the experience pool in the experience pool in S2, and adjusting the confidence in the cable abnormality detection pre-training model by using the proportion; and fine-tuning the cable abnormality detection pre-training model with adjusted confidence by using the experience pool to obtain a final cable abnormality detection model; S4, using the final cable abnormality detection model to detect the abnormality of the vibration signal generated by the to-be-detected cable in the preset length laying, and the clamp shedding position corresponding to the to-be-detected cable in the detection abnormality is determined by the similarity between the vibration signal corresponding to the to-be-detected cable in the detection abnormality and each sample in the training data set.

2. The abnormality detection method according to claim 1, characterized by, In the abnormality detection method, the corresponding clamp shedding position in the training data set is a position without setting a clamp in the preset length laying of the cable, and the label corresponding to the cable vibration signal comprises a normal fixing label and an abnormal fixing label.

3. The abnormality detection method according to claim 2, characterized by, In the abnormality detection method, when pre-training or fine-tuning the cable abnormality detection model, the parameters of the cable abnormality detection model are updated by using the following cross-entropy loss: , wherein, denotes an abnormal fixed label, denotes a cross-entropy loss, denotes the probability that the cable vibration signal is classified into the abnormal fixed label, and when fine-tuning the cable anomaly detection model, the learning rate of the cable anomaly detection model is set to 0.

001.

4. The abnormality detection method according to claim 1, characterized by, In S3, the confidence in the cable abnormality detection pre-training model is adjusted by the following method: According to the increase of the proportion of the new sample in the experience pool in the experience pool, the confidence corresponding to the probability that the cable vibration signal is classified in the abnormal fixing label is increased.

5. The abnormality detection method according to claim 4, characterized by, The confidence corresponding to the probability that the cable vibration signal is classified in the abnormal fixing label is increased by the following method: When the proportion of the new sample in the experience pool in the experience pool is greater than 5% and less than 10%, the confidence corresponding to the probability that the cable vibration signal is classified in the abnormal fixing label is 90%; When the proportion of the new sample in the experience pool in the experience pool is greater than or equal to 10%, the confidence corresponding to the probability that the cable vibration signal is classified in the abnormal fixing label is 95%.

6. The abnormality detection method according to claim 5, characterized by, In S4, the similarity between the vibration signal corresponding to the to-be-detected cable in the detection abnormality and each sample in the training data set is determined by the following method: The similarity between the vibration signal corresponding to the to-be-detected cable in the detection abnormality and the vibration signal in each sample in the training data set is calculated. The clamp falling position label corresponding to the maximum similarity among all the calculated similarities is taken as the corresponding clamp falling position of the cable to be detected when detecting an abnormality.

7. The abnormality detection method according to claim 5, characterized by, In the S4, the similarity between the vibration signal corresponding to the cable to be detected when detecting an abnormality and the vibration signal in each sample in the training data set is determined in the following manner: The similarity between the vibration signal corresponding to the cable to be detected when detecting an abnormality and the vibration signal in each sample in the training data set is calculated. All the calculated similarities are sorted from large to small, and the clamp falling position labels corresponding to the top N positions in the sorting are taken as the corresponding clamp falling positions of the cable to be detected when detecting an abnormality, wherein the N is determined according to the fine-tuning times of the cable abnormality detection model.

8. An abnormality detection system of a cable clamp characterized by comprising: The abnormality detection system comprises a vibration sensor configured to collect a vibration signal generated by a cable after being laid in a designated area. A server connected with the vibration sensor and configured to perform the steps of the method according to any one of claims 1-7.

9. A computer-readable storage medium, characterized in that, A computer program is stored thereon, and the computer program can be executed by a processor to implement the steps of the method according to any one of claims 1-7.

10. An electronic device, comprising: Comprise: One or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, cause the electronic device to implement the steps of the method according to any one of claims 1-7.

Images