Stranded cable intelligent detection method and system based on machine vision

Abstract

The invention discloses an intelligent cable stranded wire detection method and system based on machine vision, and relates to the technical field of data identification, and the method comprises the steps: carrying out the real-time image collection of a cable stranded wire through a sensor, and carrying out the preprocessing; constructing an adaptive model and a quality evaluation model based on a deep learning algorithm, and performing feature extraction on the real-time image data; inputting the extracted feature data into a quality evaluation model and comparing the extracted feature data with a preset quality standard; collecting a quality evaluation result in real time, and carrying out statistical analysis on quality data of the whole production line; and constructing a feedback mechanism, and automatically adjusting production parameters by the production line according to a quality evaluation result. According to the method, subjectivity and inaccuracy of traditional manual detection are overcome, a feedback mechanism automatically adjusts the production parameters according to the quality evaluation result, automatic control of the production process is achieved, time delay and errors caused by manual intervention and adjustment of the parameters are reduced, the production process is further optimized, and the product percent of pass is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data recognition, and in particular to a method and system for intelligent detection of cable strands based on machine vision. Background Art

[0002] In the past, manual sampling was mostly used for cable strand inspection. The inspectors checked the appearance and size of the cable strands by naked eye observation and simple measuring tools. This method is not only inefficient, but also easily affected by the subjective factors of the inspectors. Manual sampling can only cover a small number of samples and cannot fully inspect each cable strand. There are a large number of quality risks that cannot be discovered in time. Although some mechanical contact-based measuring equipment can improve the measurement accuracy to a certain extent, such equipment can usually only detect specific quality parameters, and mechanical contact detection may cause damage to the surface of the cable strands, affecting product performance. With the advancement of science and technology, machine vision technology has gradually emerged in the field of industrial inspection. The machine vision system can quickly and accurately obtain the image information of the object and analyze it through advanced algorithms. It has the advantages of non-contact, high precision and good repeatability. It has replaced the traditional detection method in many industrial production links. In the cable strand production industry, the use of machine vision technology to achieve intelligent detection has become a key direction for improving product quality and production efficiency. Through real-time image acquisition of cable strands, a large amount of appearance and structure data can be obtained, providing a rich source of information for comprehensive quality assessment.

[0003] However, the current common solutions have many shortcomings, including: the traditional manual sampling method relies on the subjective judgment of the inspectors, which is prone to visual fatigue and human errors, resulting in inaccurate test results, and manual sampling can only cover a small number of samples, and it is impossible to fully detect every cable strand, so a large number of quality problems may be missed. Traditional methods make it difficult to conduct effective statistical analysis of test data. The results of manual sampling are often just simple records, lacking systematic data analysis, and it is impossible to establish an effective feedback mechanism. The test results cannot be used in a timely manner to adjust production parameters, resulting in quality problems that persist in the production process and cannot be corrected in a timely manner, reducing product consistency and overall quality. Summary of the invention

[0004] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0005] In view of the above-mentioned problems existing in the existing intelligent detection method and system for cable strands based on machine vision, the present invention is proposed.

[0006] Therefore, the purpose of the present invention is to provide a method and system for intelligent detection of cable strands based on machine vision, which is suitable for solving the problems that traditional methods are difficult to conduct effective statistical analysis of detection data, lack of systematic data analysis, and cannot establish an effective feedback mechanism. The detection results cannot be used in time to adjust production parameters, resulting in quality problems that persist in the production process and cannot be corrected in time, thereby reducing the consistency and overall quality of the product.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] In the first aspect, an embodiment of the present invention provides an intelligent detection method for cable strands based on machine vision, which includes using sensors to capture real-time images of cable strands and perform preprocessing; building an adaptive model and a quality assessment model based on a deep learning algorithm and performing feature extraction on real-time image data; inputting the extracted feature data into the quality assessment model and comparing it with preset quality standards; collecting quality assessment results in real time, and performing statistical analysis on the quality data of the entire production line; and building a feedback mechanism so that the production line automatically adjusts production parameters according to the quality assessment results.

[0009] As a preferred solution of the cable strand intelligent detection method based on machine vision described in the present invention, wherein: the sensor is a high-resolution machine vision camera; the real-time image data includes appearance morphological data, size-related data and color information; the feature data includes geometric feature data, texture feature data, color feature data and abstract feature data.

[0010] As a preferred solution of the intelligent detection method of cable strands based on machine vision described in the present invention, the specific steps of constructing the adaptive model are: collecting a large amount of real-time image data and historical image data through a machine vision camera and annotating them; constructing an adaptive model based on a recurrent neural network RNN ​​and initializing the weights in the adaptive model; using a deep learning algorithm to extract features from real-time image data; using an optimization algorithm to update the weights of the model, and using historical image data to evaluate the performance of the model; if the evaluation result of the adaptive model is not ideal, the adaptive model needs to be optimized again.

[0011] As a preferred solution of the cable strand intelligent detection method based on machine vision of the present invention, the specific formula for extracting features of real-time image data by the deep learning algorithm is:

[0012]

[0013] Among them, Z is the comprehensive eigenvalue of the feature vector set extracted by the deep learning algorithm; x i1 is the data of the first dimension of the i-th image vector in the real-time image data set; ω1 is the first element of the adaptive model weight vector; n is the number of real-time image data.

[0014] As a preferred solution of the intelligent cable strand detection method based on machine vision described in the present invention, the specific situation of the comprehensive eigenvalue extracted by the deep learning algorithm is: when the comprehensive eigenvalue is equal to 0, it means that in the real-time image data set, the sum of the data of all image vectors in the first dimension is zero, or in the information filtering link, the sum of the proportions of the data of each image in this dimension relative to its weighted sum is zero, indicating that the relationship between the data and the weight in this dimension makes the proportion very small; when the comprehensive eigenvalue is greater than 0, it means that in the image data, the data of the first dimension has a certain cumulative amount, and also has a positive contribution after information filtering, which means that the image features represented by this dimension are reflected in the image, and after filtering the adjusted information by the weights, these features are considered to be valuable.

[0015] As a preferred solution of the cable strand intelligent detection method based on machine vision of the present invention, the specific formula of the automatic adjustment of the production parameter probability is as follows:

[0016]

[0017] Among them, P is the probability of automatically adjusting production parameters; r i is the i-th element in the quality assessment result vector; s i is the i-th element in the preset quality standard vector; m is the number of quality features.

[0018] As a preferred solution of the cable strand intelligent detection method based on machine vision described in the present invention, the specific circumstances of automatically adjusting the production parameters are as follows: when the probability of automatically adjusting the production parameters is less than a threshold value, it means that the quality assessment result is close to the preset quality standard, and the necessity of automatically adjusting the production parameters is low, and the production line does not need to adjust the production parameters; when the probability of automatically adjusting the production parameters is greater than a threshold value, it means that the difference between the quality assessment result and the preset quality standard is large, and the probability of automatically adjusting the production parameters is high, and the production line needs to automatically adjust the production parameters.

[0019] In the second aspect, in order to further solve the above technical problems, the present invention provides an intelligent detection system for cable strands based on machine vision, which includes: a data acquisition module for collecting real-time image data of cable strands during the production process; a model building module for building adaptive models and quality assessment models; a quality comparison module for comparing feature data with preset quality standards; a quality assessment module for statistically analyzing the quality data of the entire production line; and a feedback adjustment module for automatically adjusting production parameters according to the quality assessment results.

[0020] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program is executed by the processor, any step of a method for intelligent detection of cable strands based on machine vision as described in the first aspect of the present invention is implemented.

[0021] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of a method for intelligent detection of cable strands based on machine vision as described in the first aspect of the present invention.

[0022] The beneficial effects of the present invention are as follows: the present invention overcomes the subjectivity and inaccuracy of traditional manual inspection by combining image acquisition based on machine vision with feature extraction and quality assessment model of deep learning algorithm. The deep learning algorithm can automatically learn and identify complex features in cable strand images, providing detailed quality data support for the production process. The feedback mechanism automatically adjusts production parameters according to the quality assessment results, realizing automated control of the production process, which can promptly correct quality deviations caused by equipment wear, raw material changes or other factors, ensure the stability and consistency of product quality, reduce time delays and errors that may be caused by manual intervention in adjusting parameters, further optimize the production process, reduce the defective rate, and improve the product qualification rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0024] Figure 1 This is a flow chart for implementing the present invention in Example 1. DETAILED DESCRIPTION

[0025] 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.

[0026] 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.

[0027] 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.

[0028] Example 1

[0029] Reference Figure 1 , which is the first embodiment of the present invention, provides a cable strand intelligent detection method based on machine vision, comprising the following steps:

[0030] S1: Use sensors to collect real-time images of cable strands and perform preprocessing.

[0031] Preferably, Figure 1 The figure shows the implementation process of the present invention. First, the sensor is used to collect real-time images of cable strands and perform preprocessing. Then, an adaptive model and a quality assessment model are constructed based on a deep learning algorithm and features are extracted from the real-time image data. Subsequently, the extracted feature data is input into the quality assessment model and compared with the preset quality standards. The quality assessment results are collected in real time, and the quality data of the entire production line are statistically analyzed. Finally, a feedback mechanism is constructed, and the production line automatically adjusts production parameters according to the quality assessment results.

[0032] Furthermore, the sensor is a high-resolution machine vision camera.

[0033] Preferably, high-resolution machine vision cameras can obtain high-quality images of cable strands, provide rich detail information, and collect multi-dimensional data, making quality assessment more comprehensive and accurate, reducing misjudgments caused by insufficient single data.

[0034] Specifically, preprocessing mainly includes removing noise from the image, improving image quality, geometric correction and image normalization.

[0035] S2: Build adaptive models and quality assessment models based on deep learning algorithms and perform feature extraction on real-time image data.

[0036] Preferably, the real-time image data includes appearance data, size-related data and color information.

[0037] Specifically, the specific steps to build an adaptive model are:

[0038] Collect and annotate a large amount of real-time image data and historical image data through machine vision cameras;

[0039] Build an adaptive model based on the recurrent neural network RNN ​​and initialize the weights in the adaptive model;

[0040] Use deep learning algorithms to extract features from real-time image data;

[0041] Use optimization algorithms to update the model weights and use historical image data to evaluate the model performance;

[0042] If the evaluation result of the adaptive model is not ideal, the adaptive model needs to be optimized again.

[0043] Furthermore, the specific formula for deep learning algorithm to extract features from real-time image data is:

[0044]

[0045] Among them, Z is the comprehensive eigenvalue of the feature vector set extracted by the deep learning algorithm; x i1 is the data of the first dimension of the i-th image vector in the real-time image data set; ω1 is the first element of the adaptive model weight vector; n is the number of real-time image data.

[0046] Furthermore, the specific situation of the comprehensive feature value extracted by the deep learning algorithm is:

[0047] When the comprehensive eigenvalue is equal to 0, it means that in the real-time image data set, the sum of the data of all image vectors in the first dimension is zero or in the information filtering link, the sum of the proportion of the data of each image in this dimension relative to its weighted sum is zero, indicating that the relationship between the data and the weight in this dimension makes the proportion very small;

[0048] When the comprehensive eigenvalue is greater than 0, it means that in the image data, the data of the first dimension has a certain amount of accumulation and also contributes positively after information filtering, which means that the image features represented by this dimension are reflected in the image, and after filtering the adjusted information by weight, these features are considered valuable;

[0049] The larger the comprehensive eigenvalue is, the more obvious the image features are in this dimension, and the more prominent they are after information filtering. If the comprehensive eigenvalue is much larger than the values ​​calculated by other similar formulas for different dimensions, then this dimension plays a more critical role in judging the quality of cable strands.

[0050] Preferably, the adaptive model and quality assessment model constructed by the deep learning algorithm can automatically learn and extract the complex features of cable strands, improve the accuracy of detection, adapt to different types of cable strands and changes in the production environment, and have strong flexibility and adaptability.

[0051] S3: The extracted feature data are input into the quality assessment model and compared with the preset quality standards.

[0052] Preferably, when comparing the characteristic data with the preset quality standard, the quality assessment model does not perform a simple one-to-one comparison, but rather assigns weights according to the importance of different characteristics to the overall quality of the cable strands.

[0053] Specifically, in the process of comparing characteristic data with preset quality standards, there is a certain uncertainty due to sensor measurement errors, approximate calculations in data processing, and uncertainty factors in the production environment. The quality assessment model can quantify this uncertainty and reflect it in the comparison results.

[0054] S4: Collect quality assessment results in real time and perform statistical analysis on the quality data of the entire production line.

[0055] Preferably, real-time statistical analysis of quality assessment results can promptly identify quality fluctuations and trends in the production process and provide a decision-making basis for production management.

[0056] Preferably, the feedback mechanism enables the production line to automatically adjust production parameters according to the quality assessment results, ensuring the stability and consistency of product quality and reducing the defective rate.

[0057] S5: Build a feedback mechanism so that the production line automatically adjusts production parameters based on quality assessment results.

[0058] Specifically, the specific formula for automatically adjusting the probability of production parameters is as follows:

[0059]

[0060] Among them, P is the probability of automatically adjusting production parameters; r i is the i-th element in the quality assessment result vector; s i is the i-th element in the preset quality standard vector; m is the number of quality features.

[0061] Preferably, the specific conditions for automatically adjusting the production parameters are as follows:

[0062] When the probability of automatically adjusting production parameters is less than the threshold, it means that the quality assessment result is close to the preset quality standard, the necessity of automatically adjusting production parameters is low, and the production line does not need to adjust production parameters;

[0063] When the probability of automatically adjusting production parameters is greater than the threshold, it means that the quality assessment result is greatly different from the preset quality standard, the probability of automatically adjusting production parameters is high, and the production line needs to automatically adjust production parameters.

[0064] Furthermore, the threshold needs to be determined based on quality requirements, production costs, efficiency, experiments, and data analysis. By setting the threshold, unnecessary production parameter adjustments caused by data fluctuations or small deviations can be reduced, and the system can be prevented from constantly hesitating between adjustment and non-adjustment when the probability value is in an intermediate fuzzy state, thereby enhancing the certainty of decision-making.

[0065] This embodiment also provides a cable strand intelligent detection system based on machine vision, including: a data acquisition module, which is used to collect real-time image data of cable strands during the production process; a model building module, which is used to build an adaptive model and a quality assessment model; a quality comparison module, which is used to compare feature data with preset quality standards; a quality assessment module, which is used to perform statistical analysis on the quality data of the entire production line; and a feedback adjustment module, which is used to automatically adjust production parameters according to the quality assessment results.

[0066] This embodiment also provides a computer device, which is suitable for a method for intelligent detection of cable strands based on machine vision, including: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement a method for intelligent detection of cable strands based on machine vision as proposed in the above embodiment.

[0067] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0068] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, a method for intelligent detection of cable strands based on machine vision is implemented as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic storage, flash memory, disk or optical disk.

[0069] In summary, the present invention overcomes the subjectivity and inaccuracy of traditional manual inspections by combining image acquisition based on machine vision with feature extraction and quality assessment models using deep learning algorithms. The deep learning algorithm can automatically learn and identify complex features in cable strand images, providing detailed quality data support for the production process. The feedback mechanism automatically adjusts production parameters based on quality assessment results, thereby realizing automated control of the production process. This can promptly correct quality deviations caused by equipment wear, changes in raw materials or other factors, ensure the stability and consistency of product quality, reduce time delays and errors caused by manual intervention in adjusting parameters, further optimize the production process, reduce the defective rate, and improve the product qualification rate.

[0070] Example 2

[0071] Referring to Tables 1 and 2, the second embodiment of the present invention is shown. This embodiment is different from the first embodiment in that, in order to verify its beneficial effects, operating data and related instructions of the present invention in an actual environment are provided.

[0072] Tables 1 and 2 show the comparison between the detection of the present invention and the traditional detection in terms of feature extraction accuracy, quality evaluation and substandard evaluation rate in this example.

[0073] Table 1 Feature extraction accuracy comparison table

[0074]

[0075] Table 2 Comparison of quality assessment and defective rate

[0076]

[0077] As can be seen from the above table, compared with traditional detection, the present invention overcomes the subjectivity and inaccuracy of traditional manual detection. The deep learning algorithm can automatically learn and identify complex features in cable strand images, providing detailed quality data support for the production process. The feedback mechanism automatically adjusts production parameters according to the quality assessment results, realizing automated control of the production process, which can promptly correct quality deviations caused by equipment wear, raw material changes or other factors, ensure the stability and consistency of product quality, reduce the time delays and errors that may be caused by manual intervention in adjusting parameters, further optimize the production process, reduce the defective rate, and improve the product qualification rate.

[0078] 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 method for intelligent detection of cable strands based on machine vision, characterized in that: include: Use sensors to collect real-time images of cable strands and perform preprocessing; Build adaptive models and quality assessment models based on deep learning algorithms and extract features from real-time image data; The extracted feature data are input into the quality assessment model and compared with the preset quality standards; Collect quality assessment results in real time and conduct statistical analysis on quality data of the entire production line; Build a feedback mechanism so that the production line automatically adjusts production parameters based on quality assessment results.

2. The intelligent detection method for cable strands based on machine vision according to claim 1, characterized in that: The sensor is a high-resolution machine vision camera; The real-time image data includes appearance data, size-related data and color information; The feature data includes geometric feature data, texture feature data, color feature data and abstract feature data.

3. The intelligent detection method for cable strands based on machine vision according to claim 2, characterized in that: The specific steps of constructing the adaptive model are: Collect and annotate a large amount of real-time image data and historical image data through machine vision cameras; Build an adaptive model based on the recurrent neural network RNN ​​and initialize the weights in the adaptive model; Use deep learning algorithms to extract features from real-time image data; Use optimization algorithms to update the model weights and use historical image data to evaluate the model performance; If the evaluation result of the adaptive model is not ideal, the adaptive model needs to be optimized again.

4. The intelligent detection method for cable strands based on machine vision according to claim 3, characterized in that: The specific formula for the deep learning algorithm to extract features from real-time image data is: Among them, Z is the comprehensive eigenvalue of the feature vector set extracted by the deep learning algorithm; x i1 is the data of the first dimension of the i-th image vector in the real-time image data set; ω1 is the first element of the adaptive model weight vector; and n is the number of real-time image data.

5. The intelligent detection method for cable strands based on machine vision according to claim 4, characterized in that: The specific situation of the comprehensive feature value extracted by the deep learning algorithm is: When the comprehensive eigenvalue is equal to 0, it means that in the real-time image data set, the sum of the data of all image vectors in the first dimension is zero or in the information filtering link, the sum of the proportion of the data of each image in this dimension relative to its weighted sum is zero, indicating that the relationship between the data and the weight in this dimension makes the proportion very small; When the comprehensive eigenvalue is greater than 0, it means that in the image data, the data of the first dimension has a certain cumulative amount and also contributes positively after information filtering, which means that the image features represented by this dimension are reflected in the image, and after filtering the adjusted information through the weights, these features are considered valuable.

6. The intelligent detection method for cable strands based on machine vision according to claim 1, characterized in that: The specific formula for the probability of automatically adjusting production parameters is as follows: Among them, P is the probability of automatically adjusting production parameters; r i is the i-th element in the quality assessment result vector; s i is the i-th element in the preset quality standard vector; m is the number of quality features.

7. The intelligent detection method for cable strands based on machine vision according to claim 6, characterized in that: The specific conditions of automatically adjusting the production parameters are as follows: When the probability of automatically adjusting production parameters is less than the threshold, it means that the quality assessment result is close to the preset quality standard, the necessity of automatically adjusting production parameters is low, and the production line does not need to adjust production parameters; When the probability of automatically adjusting production parameters is greater than the threshold, it means that the quality assessment result is greatly different from the preset quality standard, the probability of automatically adjusting production parameters is high, and the production line needs to automatically adjust production parameters.

8. A cable strand intelligent detection system based on machine vision, based on the cable strand intelligent detection method based on machine vision according to any one of claims 1 to 7, characterized in that: include, A data acquisition module is used to collect real-time image data of cable strands during the production process; Model building module. Used to build adaptive models and quality assessment models; A quality comparison module is used to compare the characteristic data with the preset quality standards; Quality assessment module, used to perform statistical analysis on the quality data of the entire production line; Feedback adjustment module, used to automatically adjust production parameters according to quality assessment results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the cable strand intelligent detection method based on machine vision as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a cable strand intelligent detection method based on machine vision as described in any one of claims 1 to 7 are implemented.

Images