Visual inspection and sorting system for pe bottles and method of operation thereof

Abstract

The application discloses a visual detection and sorting system for PE bottles and an operation method thereof, relates to the technical field of PE bottle detection, and solves the problem that the traditional detection and sorting mode relying on manual naked-eye observation of PE bottles cannot meet the detection requirements in terms of efficiency and precision, and visual fatigue is prone to occur in the case of long-time work of employees or mass production, thereby leading to missing and misjudgment, causing unqualified PE bottles to flow into the market, and bringing negative effects to the image of enterprises. The visual detection and sorting system for PE bottles combines image processing, machine learning and automation technology, and can realize rapid and accurate detection and sorting of the quality of PE bottles. It can not only greatly improve the efficiency and precision of PE bottle detection and reduce manual visual fatigue, but also can avoid unqualified products from flowing into the market, and guarantee the image of enterprises and the rights and interests of consumers.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of PE bottle detection, specifically to a visual detection and sorting system for PE bottles and an operation method thereof. BACKGROUND

[0002] PE bottles are common packaging materials for filling beverages, cosmetics, cleaning agents, and other products. Quality problems can cause PE bottles to have poor sealing, easy to leak or deform, affecting the quality and use experience of products. Through detection, it can ensure that the appearance of PE bottles is intact, the size is accurate, and the presence of defects or contaminants in PE bottles is excluded, improving the quality and reliability of products, and ensuring the safety and reliability of PE bottles.

[0003] Traditional PE bottle quality detection and sorting relies on manual visual observation, and employees check on the assembly line. However, this method cannot meet the detection standards in terms of efficiency, accuracy, and employee visual fatigue. Especially in the case of mass production, there are often omissions and misjudgments, causing some substandard bottled beverages to flow into the market, negatively affecting the image of the enterprise.

[0004] Therefore, the present application provides a visual detection and sorting system for PE bottles and an operation method thereof to solve the above problems. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a visual detection and sorting system for PE bottles and an operation method thereof, which solves the problem that the traditional method of relying on manual visual observation to detect and sort PE bottles cannot meet the detection requirements in terms of efficiency, accuracy, and the like. Employees are prone to visual fatigue during long-term work or mass production, resulting in omissions and misjudgments, causing substandard PE bottles to flow into the market, and negatively affecting the image of the enterprise.

[0006] To achieve the above purpose, the present application realizes the following technical solutions: a visual detection and sorting system for PE bottles, comprising:

[0007] A central processing module for controlling the operation and operation of the entire system;

[0008] A feeding unit for introducing PE bottles to be detected and sorted into the system;

[0009] An image acquisition unit for acquiring images of PE bottles and transmitting image data to a subsequent image processing module;

[0010] An image processing unit responsible for processing and analyzing the PE bottle images collected by the image acquisition unit;

[0011] Feature classification unit: input the extracted features into machine learning or deep learning algorithms for classification;

[0012] Sorting unit: according to the detection results of the image processing module, the sorting system will classify and sort the PE bottles;

[0013] Data recording and statistical module: used to record detection results and statistical data, including the number of qualified PE bottles, the number of unqualified PE bottles, and the error rate;

[0014] Data feedback unit: the system collects sorting results, error discrimination rate and other data, and performs statistical analysis, and according to the analysis results, optimizes and improves the image processing and classification algorithm, improves the accuracy and performance of the system.

[0015] Further, the feeding unit includes a conveyor belt, a transport mechanism, a sensor, a positioning device and a track / guide device.

[0016] Further, the feature classification unit includes a feature extractor, a feature selector, a feature encoder, a feature selection and feature encoding evaluator.

[0017] Further, the sorting unit includes a conveying system, a sorting area, an identification system, a sorting algorithm and a sorting device.

[0018] The present application also provides an operating method for the visual detection and sorting system of PE bottles, which comprises the following steps:

[0019] Step one, start the system: the operator first starts the control interface of the system, including connecting the image acquisition unit and other equipment controllers, ensuring the connection with the related equipment is normal, the system enters standby state, preparing to receive the image from the image acquisition unit;

[0020] Step two, set parameters: the operator sets the detection parameters and sorting method of PE bottles according to specific requirements on the control interface;

[0021] Step three, prepare PE bottles: place the PE bottles to be detected and sorted in the feeding system, ensure that the PE bottles can smoothly enter the system for detection and sorting;

[0022] Step four, start detection: the operator clicks the start detection button on the control interface, the system starts to automatically detect and sort the PE bottles, captures the image of the PE bottles in real time through the image acquisition unit, and transmits the image to the image processing module for processing;

[0023] Step five, image processing: the image processing module uses the algorithm set in advance to preprocess and analyze the PE bottle image, and extract the features of the PE bottle;

[0024] Step six, feature classification: the extracted PE bottle features are input into a deep learning algorithm for classification to determine the quality and defects of the PE bottle.

[0025] Step seven, sorting operation: according to the classification results, the central processing module will remove the unqualified PE bottles from the production line, and send the qualified PE bottles to the next packaging and delivery link;

[0026] Step eight, result recording and statistics: the data recording and statistics module automatically records the collection and sorting results of each batch, error discrimination rate and other data, and the operator views the statistical data through the control interface and performs subsequent data analysis and quality control;

[0027] Step nine, data feedback and optimization: the system collects sorting results, error discrimination rate and other data, and performs statistics and analysis, and according to the analysis results, the image processing and classification algorithm is optimized and improved to improve the accuracy and performance of the system.

[0028] Further, in step two, the detection parameters of the PE bottle include the size, shape, color and identifier of the PE bottle.

[0029] Further, in step six, the deep learning algorithm has the ability to automatically determine the quality and defects of the PE bottle according to the features through a large amount of data training, and the classification results include two categories of qualified and unqualified.

[0030] Further, in step eight, the detection results and statistical data include the number of qualified PE bottles, the number of unqualified PE bottles, and the misjudgment rate.

[0031] Advantages

[0032] The present application provides a visual detection and sorting system for PE bottles and its operation method. Compared with the prior art, it has the following advantages:

[0033] 1. The visual detection and sorting system for PE bottles and its operation method, through steps one, starting the system, step two, setting parameters, step three, preparing PE bottles, step four, starting detection, step five, image processing, step six, feature classification, step seven, sorting operation, step eight, result recording and statistics, step nine, data feedback and optimization, solves the problem that the traditional manual visual observation of PE bottles for detection and sorting cannot meet the detection requirements in terms of efficiency, accuracy, etc. Visual fatigue is easy to occur in the case of long-time work of employees or large-scale production, which leads to omission and misjudgment, causing unqualified PE bottles to flow into the market, and negatively affecting the image of the enterprise.

[0034] 2、The visual detection and sorting system for PE bottles and its operation method, by combining image processing, machine learning and automation technology, can realize rapid and accurate detection and sorting of PE bottle quality. It not only can greatly improve the efficiency and accuracy of PE bottle detection, reduce manual visual fatigue, but also can avoid unqualified products flowing into the market, protect the image of enterprises and consumer rights and interests.

[0035] 3、The visual detection and sorting system for PE bottles and its operation method, through image acquisition, preprocessing, feature extraction and classification steps, can automatically detect the quality and defects of PE bottles and realize sorting operation, thereby improving the detection and sorting efficiency and accuracy of PE bottle production line. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 The structural principle diagram of the present application;

[0037] Figure 2 The structural principle diagram of the feed unit of the present application;

[0038] Figure 3 The structural principle diagram of the feature classification unit of the present application;

[0039] Figure 4 The structural principle diagram of the sorting unit of the present application;

[0040] Figure 5 The production process flow chart of the present application. DETAILED DESCRIPTION

[0041] The technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0042] Please refer to Figure 1 The embodiments of the present application provide a technical solution: a visual detection and sorting system for PE bottles, specifically including the following embodiments:

[0043] The visual detection and sorting system for PE bottles includes:

[0044] Central processing module: used for controlling the operation and operation of the whole system, the central processing module is usually composed of a computer or an embedded controller, interacts with the operator through a control interface, and controls the actions of each part. The central processing module can also set the detection parameters and sorting method of PE bottles, record the data of the detection and sorting process, etc.

[0045] Feed unit: used to introduce PE bottles to be detected and sorted into the system, which can use conveyor belts, mechanical arms or other means for transportation and positioning of PE bottles. The feeding system needs to be able to accurately deliver PE bottles to the subsequent detection and sorting modules;

[0046] Image acquisition unit: used to collect images of PE bottles and transmit image data to the subsequent image processing module. The camera system usually selects high-resolution and high-frame-rate industrial cameras to obtain clear and accurate images of PE bottles. The camera system needs to have good image acquisition capability and image transmission stability;

[0047] Image processing unit: responsible for processing and analyzing the images of PE bottles collected by the image acquisition unit. The image processing module usually uses image processing algorithms such as edge detection, color analysis, shape matching, etc. to extract and analyze the features of PE bottles. Through image processing, the quality, defects and identifiers of PE bottles can be detected;

[0048] Feature classification unit: input the extracted features into machine learning or deep learning algorithms for classification. These algorithms have been trained through a large amount of data and can automatically determine the quality and defects of PE bottles according to the features. The classification results can include two categories: qualified and unqualified;

[0049] Sorting unit: according to the detection results of the image processing module, the sorting system will classify and sort PE bottles. The sorting system can use gas injection, mechanical arms, conveyor belts or other mechanical equipment to separate qualified and unqualified PE bottles to ensure product quality;

[0050] Data recording and statistical module: used to record detection results and statistical data, including the number of qualified PE bottles, the number of unqualified PE bottles, and the error rate. The data recording and statistical module can save data to a database or generate reports, facilitating subsequent data analysis and quality control;

[0051] Data feedback unit: the system collects sorting results, error discrimination rate and other data, and performs statistical analysis. According to the analysis results, the image processing and classification algorithms are optimized and improved to improve the accuracy and performance of the system.

[0052] The image acquisition unit includes a camera: the camera is the core component of the image acquisition device, used to capture the image of the object to be detected. The camera usually includes an image sensor and a lens. In the case of sufficient light, it will convert the optical information of the object into an electrical signal to form an image;

[0053] Synchronous flash: Synchronous flash is used to provide additional light source to improve the brightness and contrast of the image. In some cases, when the light is insufficient, the synchronous flash can be used to supplement the light, thereby improving the quality of the image;

[0054] Image processor: Image processor is used to process and handle the collected images. This includes image filtering, edge detection, brightness adjustment, etc. Through the image processor, the image quality can be optimized, and specific image features can be extracted for subsequent analysis and judgment;

[0055] Image transmission interface: Image transmission interface is used to transmit the collected images to other systems or devices for further processing. Common image transmission interfaces include USB, Ethernet, HDMI, etc.

[0056] Data collection: Data collection refers to collecting raw data from various sources such as sensors, questionnaires, web crawlers, etc. Data collection can be real-time or periodic, and can be manually collected or automatically collected;

[0057] Data cleaning and preprocessing: Data cleaning and preprocessing is to filter, clean, denoise, remove duplicates, fill missing values, etc. to ensure the quality and consistency of the data. This step usually involves data cleaning, data validation and data reformatting operations. Data storage: Data storage is to save the cleaned and preprocessed data into a database or file for subsequent use and analysis. Common data storage methods include relational databases, non-relational databases, data warehouses, data lakes, etc.

[0058] Data analysis: Data analysis is to further mine and analyze the stored data. Through the use of different statistical methods, machine learning algorithms and visualization tools, patterns, trends, correlations and other information can be discovered in the data, and meaningful results can be extracted;

[0059] Data visualization: Data visualization is to display the analyzed data in the form of charts, graphs, maps, etc. to make it easier for people to observe and analyze the data more intuitively and easily. Data visualization can be achieved through the use of various visualization tools and programming languages;

[0060] Data reporting and statistical analysis: Data reporting and statistical analysis is to organize the results of data analysis and visualization into reports, slides, documents, etc. to facilitate the communication, sharing and interpretation of data results to stakeholders. Statistical analysis can include descriptive statistics, inferential statistics, etc.

[0061] In the embodiments of the present application, the feeding unit includes a conveyor belt: the conveyor belt is the main component of the feeding system, which is used to transport the PE bottles to be detected to the detection and sorting area. The conveyor belt is usually composed of one or more devices with protrusions, which are used to push the PE bottles to the next processing step. Transport mechanism: the transport mechanism is used to transport the PE bottles from the starting position to the detection and sorting area. This can include a conveying chain, a conveying belt, air flow conveying, etc. The transport mechanism needs to have stable and reliable functions to ensure that the PE bottles can accurately and quickly reach the detection area. Sensor: the feeding system usually uses sensors to detect the arrival of the PE bottles and confirm the status of the conveyor belt. For example, using a photoelectric sensor to detect the presence or absence of PE bottles, or using a weighing sensor to detect the weight of the PE bottles. The sensor can be used to trigger image acquisition and other next-step processing operations. Positioning device: the positioning device is used to ensure the accurate position of the PE bottles in the detection and sorting area. It can accurately position the PE bottles during the conveying process through mechanical devices, electromagnetic devices or pneumatic devices, etc. so that the subsequent detection and sorting operations can be carried out smoothly. Track / guide device: through suitable tracks or guide devices, the feeding system can guide the PE bottles to the correct position for subsequent detection and sorting operations. The track and guide device should ensure the smooth movement of the PE bottles and minimize the collision and extrusion between the PE bottles.

[0062] In the embodiments of the present application, the feature classification unit includes a feature extractor: the feature extractor is an algorithm or model used to extract useful information from raw data. In image classification, common feature extractors include convolutional neural networks (CNN), local binary patterns (LBP), and histograms of oriented gradients (HOG), etc. These feature extractors can convert images into corresponding feature vectors. Feature selector: the feature selector is used to select the most important or representative features from the extracted features. The purpose of the feature selector is to reduce the dimension of the feature vector and retain the most critical information to improve the accuracy and efficiency of classification. Common feature selection algorithms include variance selection method, correlation coefficient method and principal component analysis (PCA), etc. Feature encoder: the feature encoder is an algorithm that encodes or compresses the original features. Common feature encoders include Bag of Words (BoW), Autoencoder and Hash function, etc. The feature encoder can convert the original features into a more compact and reliable representation. Feature selection and feature encoding evaluator: the feature selection and feature encoding evaluator is an algorithm or index used to evaluate various feature selection and feature encoding methods. Common feature selection and feature encoding evaluation indexes include information gain, mutual information, importance evaluation and reconstruction error, etc. These evaluators are used to evaluate the effect and quality of feature selection and feature encoding. Feature classification is composed of feature extractor, feature selector, feature encoder and feature selection and feature encoding evaluator, etc. These parts work together to extract useful information from raw data and convert it into a more representative feature representation to improve the accuracy and efficiency of the classification algorithm.

[0063] In the embodiments of the present application, the sorting unit includes a conveying system: the conveying system is used to send the articles or goods to be sorted from the production line or storage area to the sorting area, and usually includes a conveyor belt, a conveying line, a chute, a conveying vehicle and the like. The sorting area: the sorting area is the area where the articles or goods to be sorted are sorted. The sorting area is usually provided with a sorting table, a shelf, a storage box or a sorting grid and the like for temporarily storing the articles to be sorted. The identification system: the identification system is used to identify and obtain the relevant information of the articles to be sorted, such as bar code, two-dimensional code, RFID tag and the like. The identification system is usually composed of a scanning gun, a reader, a camera and the like, which can quickly read the information of the articles and input it into the sorting system. The sorting algorithm: the sorting algorithm is the core part of the sorting operation, which is used to determine where to send the articles according to the information of the articles to be sorted and the sorting rules. The sorting algorithm can be realized by rules, logic or machine learning and the like to realize the automatic sorting of the articles. The sorting device: the sorting device is a device for sending the articles to be sorted to their target positions, and common sorting devices include a conveyor belt, a warehouse robot, a sorting mechanical arm and the like. The sorting device sends the articles to the corresponding storage area, shelf or outlet according to the indication of the sorting algorithm. These parts work together to realize the fast and accurate sorting of the articles in an automated manner.

[0064] The embodiments of the present application also provide an operation method of the visual detection and sorting system for PE bottles, which comprises the following steps:

[0065] Step one, starting the system: the operator first starts the control interface of the system, including connecting the image acquisition unit and the controller of other devices, ensuring that the connection with the related devices is normal, the system enters the standby state, and is ready to receive the images transmitted by the image acquisition unit;

[0066] Step two, setting parameters: the operator sets the detection parameters and sorting method of the PE bottles on the control interface according to specific requirements;

[0067] Step three, preparing PE bottles: placing the PE bottles to be detected and sorted in the feeding system to ensure that the PE bottles can smoothly enter the system for detection and sorting, and the operator needs to correctly place the PE bottles according to the requirements to ensure the normal operation of the system;

[0068] Step four, starting detection: the operator clicks the start detection button on the control interface, and the system starts to automatically detect and sort the PE bottles, captures the images of the PE bottles in real time through the image acquisition unit, and transmits the images to the image processing module for processing;

[0069] Step five, image processing: the image processing module uses a pre-set algorithm to pre-process and analyze the PE bottle image, through edge detection, color analysis and other technologies, pre-processing includes denoising, enhancement and filtering and other operations to improve image quality and reduce interference, these operations can be realized through computer vision algorithms, and the characteristics of the PE bottle are extracted;

[0070] Step six, feature classification: the extracted PE bottle features are sent to machine learning or deep learning algorithms for classification to determine the quality and defects of the PE bottle, the algorithm has been trained through a large amount of data and has the ability to automatically determine the quality and defects of the PE bottle according to the characteristics, and the classification results include two categories of qualified and unqualified;

[0071] Step seven, sorting operation: according to the classification results, the central processing module will remove the unqualified PE bottles from the production line, and the qualified PE bottles will be sent to the next packaging and delivery link, and the sorting unit is realized through gas injection, mechanical arm or conveyor belt and other ways;

[0072] Step eight, result recording and statistics: the data recording and statistics module automatically records the collection and sorting results of each batch, error discrimination rate and other data, and the operator can view the statistical data through the control interface and perform subsequent data analysis and quality control;

[0073] Step nine, data feedback and optimization: the system collects sorting results, error discrimination rate and other data, and performs statistics and analysis, and according to the analysis results, the image processing and classification algorithm are optimized and improved to improve the accuracy and performance of the system.

[0074] In the step two of the embodiment of the application, the detection parameters of the PE bottle include the size, shape, color and identifier of the PE bottle.

[0075] In the step six of the embodiment of the application, the deep learning algorithm is trained through a large amount of data and has the ability to automatically determine the quality and defects of the PE bottle according to the characteristics, and the classification results include two categories of qualified and unqualified.

[0076] In the step eight of the embodiment of the application, the detection results and statistical data include the number of qualified PE bottles, the number of unqualified PE bottles and the misjudgment rate.

[0077] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting; it is not intended to exclude myriad other embodiments of the present application that other inventors can develop based on the same general inventive concepts embodied by the described embodiments. That is, although the present application is described in terms of particular embodiments and illustrative figures, it should be apparent that the scope of the present application is not limited to these specific embodiments.

[0078] While the embodiments of the application have been shown and described herein, it will be understood by those skilled in the art that many changes, modifications, substitutions and alterations to these embodiments can be made without departing from the principles and spirits of the application, and it is intended that the scope of the application be limited solely by the scope of the appended claims and the equivalents thereof.

Claims

1. A vision inspection and sorting system for PE bottles, characterized by: The system comprises: a central processing module for controlling the operation of the entire system; a feeding unit for introducing PE bottles to be detected and sorted into the system; an image acquisition unit for acquiring images of the PE bottles and transmitting image data to the subsequent image processing module; an image processing unit responsible for processing and analyzing the images of the PE bottles collected by the image acquisition unit; a feature classification unit that inputs the extracted features into a machine learning or deep learning algorithm for classification; a sorting unit that sorts the PE bottles according to the detection results of the image processing module; a data recording and statistical module for recording detection results and statistical data, including the number of qualified PE bottles, the number of unqualified PE bottles, and the error rate; a data feedback unit that collects sorting results, error discrimination rates, and other data, and performs statistical analysis and optimization of image processing and classification algorithms to improve the accuracy and performance of the system.

2. The visual inspection and sorting system for PE bottles of claim 1, characterized in that: The feeding unit includes a conveyor belt, a transport mechanism, sensors, positioning devices, and tracks / guides.

3. The visual inspection and sorting system for PE bottles of claim 1, wherein: The feature classification unit includes a feature extractor, a feature selector, a feature encoder, a feature selection and feature encoding evaluator.

4. The visual inspection and sorting system for PE bottles of claim 1, wherein: The sorting unit includes a conveying system, a sorting area, an identification system, a sorting algorithm, and sorting equipment.

5. A method of operating a vision inspection and sorting system for PE bottles, characterized by: The visual detection and sorting system for PE bottles according to any one of claims 1-4, specifically comprising the following steps: Step one, start the system: the operator first starts the control interface of the system, including connecting the image acquisition unit and other equipment controllers, ensuring that the connection with the related equipment is normal, and the system enters the standby state, preparing to receive the images transmitted by the image acquisition unit; Step two, set parameters: the operator sets the detection parameters and sorting method of the PE bottles on the control interface according to specific requirements; Step three, prepare PE bottles: place the PE bottles to be detected and sorted in the feeding system to ensure that the PE bottles can smoothly enter the system for detection and sorting; Step four, start detection: the operator clicks the start detection button on the control interface, and the system starts automatic detection and sorting of the PE bottles, captures the images of the PE bottles in real time through the image acquisition unit, and transmits the images to the image processing module for processing; Step five, image processing: the image processing module uses the pre-set algorithm to preprocess and analyze the PE bottle images, extract the features of the PE bottles, and judge the quality and defects of the PE bottles; Step six, feature classification: the extracted features of the PE bottles are input into the deep learning algorithm for classification; Step seven, sorting operation: according to the classification results, the central processing module will remove the unqualified PE bottles from the production line, and send the qualified PE bottles to the next packaging and shipping link; Step eight, result recording and statistics: the data recording and statistical module automatically records the collection and sorting results, error discrimination rates, and other data of each batch, and the operator views the statistical data through the control interface and performs subsequent data analysis and quality control. Step nine, data feedback and optimization: the system collects data such as sorting results and error discrimination rate, and conducts statistics and analysis, and according to the analysis results, the image processing and classification algorithm are optimized and improved to improve the accuracy and performance of the system.

6. The method of operating a visual inspection and sorting system for PE bottles according to claim 5, characterized in that: In the second step, the detection parameters of the PE bottle include the size, shape, color and identifier of the PE bottle.

7. The method of operating a visual inspection and sorting system for PE bottles of claim 5, wherein: In the sixth step, the deep learning algorithm has the ability to automatically distinguish the quality and defects of the PE bottle according to the characteristics through a large amount of data training, and the classification results include two categories of qualified and unqualified.

8. The method of operating a visual inspection and sorting system for PE bottles of claim 5, wherein: In the eighth step, the detection results and statistical data include the number of qualified PE bottles, the number of unqualified PE bottles, and the misjudgment rate.

Images