Automatic detection system and detection method for pocking marks on surfaces of plastic master batches

Abstract

The invention relates to the technical field of master batch production quality detection, in particular to an automatic detection system and method for plastic master batch surface pocking marks, and the detection system comprises a sheet forming device, a sheet conveying device, a visual detection system and a data processing system. The sheet forming device forms plastic master batches according to set temperature and set pressure to obtain sheets, the sheets are continuously conveyed to the visual detection system by the sheet conveying device according to set spacing and set speed, and the visual detection system conducts visual detection on the sheets one by one. And an image detected by the visual detection system is analyzed and processed by the data processing system to obtain the characteristic parameters and the spatial position of the pocking mark defect area. According to the invention, real-time, continuous and online monitoring of the plastic master batch production process can be realized, the hysteresis of detection is effectively eliminated, the detection efficiency and the automation degree are improved, and the detection precision and the detection reliability are improved at the same time.

Description

Technical Field

[0001] This invention relates to the technical field of masterbatch production quality inspection, and more specifically, to an automatic detection system and method for detecting surface pitting on plastic masterbatch. Background Technology

[0002] As a key raw material in plastic product processing, the appearance quality of plastic masterbatch directly affects the performance and quality of the final product. Among various opaque masterbatch products, surface pitting is a common appearance defect, typically manifesting as tiny depressions or impurities on the particle surface. These pits not only affect the aesthetics of the product but can also become stress concentration points within the material, thus weakening its mechanical properties. For PVC (polyvinyl chloride) masterbatch, widely used in wire and cable sheathing, building materials, and films, the requirements for surface quality are particularly stringent; especially in the production process of PVC cable materials, the detection of pitting is crucial. If pitting exists on the surface of the cable material, during subsequent extrusion processing, the defect location can easily lead to uneven thickness of the insulation or sheath layer, becoming a weak point under the influence of an electric field, triggering localized electrical breakdown accidents, seriously threatening the safety and lifespan of the cable, and ultimately severely impacting the quality of the end product.

[0003] Currently, the mainstream testing method for quality control of surface pitting on PVC masterbatch in the industry follows the traditional offline, destructive testing process: first, samples are taken periodically on the production line to collect a small amount of masterbatch samples, and then they are sent to the laboratory for analysis; in the laboratory, the operators rely entirely on visual observation or simple magnification equipment to manually identify and count the number of pitting on the surface of the masterbatch and assess its approximate size. This traditional method of detecting pinholes has a series of technical problems that urgently need to be solved: On the one hand, the entire process, from sampling, sample delivery, laboratory preparation to final manual analysis and results, often takes up to four hours or even longer. The severe lag in test results means that quality feedback information is far behind the actual production rhythm, resulting in the production line being in a state of "blind production" without effective quality monitoring for most of the time. This cannot meet the precise quality management requirements of modern manufacturing enterprises for real-time monitoring and immediate intervention in the production process. Due to the delay in quality information feedback, when the test results finally show quality abnormalities, a large number of unqualified products have already been produced, which can easily lead to defective products flowing into downstream customers, thereby triggering customer doubts and complaints about the quality of the entire batch of materials, causing huge economic losses and long-term brand reputation risks to the manufacturing enterprise. On the other hand, this test method is labor-intensive and inefficient. At the same time, over-reliance on human experience also brings the risk of unstable interpretation results and poor repeatability. Different operators or the same operator at different times may give different conclusions for the same sample. Summary of the Invention

[0004] The purpose of this invention is to overcome the problems of delayed detection results, unstable interpretation results leading to poor repeatability, high labor costs and low efficiency in the existing automatic detection system for surface pitting of plastic masterbatch. The invention provides an automatic detection system for surface pitting of plastic masterbatch that eliminates detection lag in real-time online detection, improves detection efficiency and automation, and enhances detection accuracy and reliability.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0006] An automatic detection system for surface pitting in plastic masterbatch is provided, comprising a sheet forming device, a sheet conveying device, a vision inspection system, and a data processing system. The sheet forming device is a screw extruder with a slit-shaped die, used to extrude the plastic masterbatch into sheets of a target size. The sheet forming device is located at one end of the sheet conveying device. The sheet conveying device includes a servo motor with a built-in encoder and a conveyor belt driven by the servo motor at a set speed. The conveyor belt receives the sheets extruded by the sheet forming device and conveys multiple sheets to the vision inspection system at a set interval and a set speed. The visual inspection system is located at the other end of the sheet transport device. It includes an image acquisition component and an image acquisition trigger component connected by a signal. The image acquisition trigger component is connected to the encoder by a signal and is used to send a signal to the image acquisition component to perform line scanning. The image acquisition component is located above the conveyor belt and is used to acquire sheet images. The direction of the line scan is perpendicular to the transport direction of the sheet. The data processing system is used to receive the sheet images and perform analysis to obtain the feature parameters of the pit defect area and the spatial location of the pit defect area. The feature parameters include area, perimeter, and roundness.

[0007] The automatic detection system for surface pitting of plastic masterbatch of the present invention can be directly integrated into the plastic masterbatch production line. The plastic masterbatch produced by the plastic masterbatch production line is quantitatively and online fed into the sheet forming device. The sheet forming device forms the plastic masterbatch into sheets according to a set temperature and a set pressure. The sheets are continuously transported to the vision inspection system by the sheet conveying device at a set interval and a set speed. The vision inspection system performs visual inspection on each sheet. The images detected by the vision inspection system are analyzed and processed by the data processing system to obtain the characteristic parameters of the pitting defect area and the spatial location of the pitting defect area.

[0008] The sheet conveying device of this invention includes a servo motor and a conveyor belt. The servo motor has an encoder that monitors the rotational speed in real time. The servo motor adjusts its operating state in real time according to the rotational speed monitored by the encoder, ensuring that the sheet is conveyed at a uniform speed and stably on the conveyor belt. This avoids imaging deviations caused by speed fluctuations or vibrations, thus helping the vision inspection system achieve high-quality imaging. In addition, the encoder outputs pulse signals in real time. The sheet conveying speed can be calculated based on the pulse signals, and the sheet conveying speed signal is converted into a camera trigger signal to control the image acquisition component to acquire images in an equidistant scanning manner, further contributing to the high-quality imaging of the vision inspection system.

[0009] The automatic detection system for surface pitting on plastic masterbatch of this invention enables real-time, continuous, and online monitoring of the plastic masterbatch production process. On the one hand, it solves the problem of traditional laboratory manual testing requiring 4 hours and the production line being in a state of blind production without monitoring, avoiding the continuous production of defective products and their flow to customers, significantly reducing quality risks and economic losses. On the other hand, the detection results are more accurate, objective, and repeatable, reducing the rate of missed and false detections, avoiding the problem of excessive reliance on operator experience, and solving the problems of subjective errors and fatigue errors that may exist in manually distinguishing the number and size of pitting, providing a reliable basis for quality judgment, improving product consistency and customer trust. Furthermore, the entire process of detecting surface pitting on plastic masterbatch is carried out automatically without manual intervention. Operators only need to pay attention to alarms and handle abnormalities, effectively improving overall production efficiency.

[0010] Preferably, the image acquisition component includes an illumination device and an imaging device located on both sides of a vertical line: the central axis of the illumination device forms an angle with the vertical line. The central axis of the imaging device is at an angle to the vertical line. ;in, , The lighting and imaging devices are both at an angle to the vertical direction, thus forming a reflective imaging architecture, which makes the pitting defects appear as obvious grayscale differences in the image.

[0011] Preferably, the device further includes a mounting bracket, a first fixed frame, and a second fixed frame. The mounting bracket has a first mounting groove and a second mounting groove, with the second mounting groove being higher than the first mounting groove. The lighting device is slidably mounted in the first mounting groove via the first fixed frame, and the imaging device is slidably mounted in the second mounting groove via the second fixed frame. Both the lighting device and the imaging device are located above the conveyor belt. The position of the lighting device is adjustable to accommodate imaging of sheets of different specifications, and the height of the imaging device is adjustable to accommodate imaging of sheets of different thicknesses, making it widely applicable.

[0012] Preferably, both the first and second fixed frames include a fixed crossbar, a fixed buckle, and a limiting piece: the end of the fixed crossbar is slidably installed in the first or second mounting groove; the limiting piece includes a first connecting part fixedly connected to the fixed crossbar, a second connecting part fixedly connected to the lighting device or imaging device, and a third connecting part disposed between the first and second connecting parts, with an included angle θ formed between the second and third connecting parts. The included angle θ is specifically the angle between the center lines of the second connecting part and the third connecting part; the fixing buckle is fixedly connected to the second connecting part and the fixing buckle is fixedly connected to the lighting device or imaging device. By setting the angles of the second and third connecting parts of the limiting piece, a reflective imaging architecture is formed. The first fixing frame, the second fixing frame, and the fixing bracket form a structurally stable installation system, thereby effectively ensuring the installation accuracy and positional stability of the lighting device and the imaging device.

[0013] Preferably, the data processing system includes an image preprocessing submodule, a feature extraction and segmentation submodule, and a defect determination submodule. The image preprocessing submodule includes a region of interest (ROI) extraction module and a noise suppression module. The ROI extraction module processes the sheet image and extracts a first image, while the noise suppression module filters noise from the first image to obtain a second image that retains the details of the pitted edges. The feature extraction and segmentation submodule processes the second image and separates the pitted defect area from the background, and connects broken pitted defect areas to restore the true shape of the pits. The defect determination submodule analyzes the feature parameters of the pitted defect area and compares these parameters with a preset threshold to classify the defect. By preprocessing the sheet image and extracting accurate and clear images of the pitted defect area through feature extraction and segmentation, and then performing defect determination and data storage, the detection efficiency and automation level can be effectively improved.

[0014] Preferably, the system also includes a data storage module and an alarm module. The data storage module records and stores the spatial coordinates and characteristic parameters of the pitting defect area, while the alarm module issues an alarm when the number of pits within a set length exceeds a set threshold. The entire pitting detection process is automated, requiring no manual intervention, thus freeing up manpower. Operators only need to focus on alarms and handling anomalies, effectively improving overall production efficiency.

[0015] Preferably, a plurality of air-cooling devices for cooling the sheet surface are evenly distributed above the sheet conveying device. The air-cooling devices cool the sheet surface and stabilize the sheet surface temperature below 100°C, avoiding the impact of high sheet surface temperature on detection accuracy.

[0016] Preferably, it also includes a cleaning and protective structure, with at least the sheet conveying device and the vision inspection system located inside the cleaning and protective structure. The cleaning and protective structure creates a relatively enclosed inspection space, reducing interference from external light and dust.

[0017] This invention also provides an automatic detection method for surface pitting in plastic masterbatch, comprising the following steps: The steps of sheet forming are as follows: plastic masterbatch is added into the screw extruder and extruded by the screw extruder to obtain a sheet of the target size; The sheet conveying process involves the sheet falling onto a conveyor belt, and the conveyor belt transporting multiple sheets to the vision inspection system at a set interval and a set speed. The steps of sheet imaging are as follows: The image acquisition trigger component receives the signal from the encoder and triggers the image acquisition component to perform line scanning on the sheet surface to obtain a sheet image; The data processing steps are as follows: Analyze and process the sheet image to obtain the characteristic parameters of the pitted defect area and the spatial location of the pitted defect area.

[0018] The automatic detection method for surface pitting on plastic masterbatch of the present invention enables real-time, continuous, and online monitoring of the plastic masterbatch production process. On the one hand, it solves the problem of traditional laboratory manual testing requiring 4 hours and the production line being in a state of blind production without monitoring, avoiding the continuous production of defective products and their flow to customers, and significantly reducing quality risks and economic losses. On the other hand, the detection results are more accurate, objective, and repeatable, reducing the rate of missed detections and false detections, avoiding the problem of excessive reliance on operator experience, and solving the problems of subjective errors and fatigue errors that may exist in manually distinguishing the number and size of pitting, providing a reliable basis for quality judgment, improving product consistency and customer trust. Furthermore, the entire process of detecting surface pitting on plastic masterbatch is carried out automatically without manual intervention. Operators only need to pay attention to alarms and handle abnormalities, effectively improving overall production efficiency.

[0019] Preferably, the data processing steps include: an image preprocessing step: determining the edge of the sheet material using an adaptive threshold segmentation algorithm, and removing the background area using Blob analysis to extract a first image; using a nonlocal mean filtering algorithm to filter salt-and-pepper noise and Gaussian noise in the region of interest image, preserving the details of the pitted edges to obtain a second image; a feature extraction and segmentation step: calculating the local standard deviation and gradient magnitude of the second image to construct a texture feature map, using a region growing algorithm to binarize the map, and separating the pitted defects from the background; removing isolated noise points using morphological opening and closing operations, connecting broken pitted defect areas, and restoring the true shape of the pitted defects; a defect determination step: using connected component analysis technology to extract feature parameters of each pitted defect area, and comparing the feature parameters with a preset determination threshold to identify and classify the pitted defect areas.

[0020] Compared with the prior art, the beneficial effects of the present invention are: The automatic detection system and method for surface pitting of plastic masterbatch of the present invention can realize real-time, continuous and online monitoring of the plastic masterbatch production process. It can solve the problems of traditional laboratory manual testing requiring 4 hours and the production line being in a state of blind production without monitoring, avoid the continuous production of defective products and their flow into customers, and significantly reduce quality risks and economic losses. The present invention discloses an automatic detection system and method for surface pitting of plastic masterbatch. The entire process of detecting surface pitting of plastic masterbatch is automated without human intervention. On the one hand, it can effectively improve overall production efficiency; on the other hand, the detection results are more accurate, objective, and repeatable, reducing the rate of missed detections and false detections. It avoids the problem of relying too much on the operator's experience and solves the problems of subjective error and fatigue error that may exist in manually distinguishing the number and size of pitting. It provides a reliable basis for quality judgment and improves product consistency and customer trust. Attached Figure Description

[0021] Figure 1 A schematic diagram of an automatic detection system for surface pitting in plastic masterbatch; Figure 2 A schematic diagram showing the relative positions of the lighting device and the imaging device; Figure 3 A schematic diagram of the principle of an automatic detection system for surface pitting in plastic masterbatch; Figure 4 Another schematic diagram of the principle of an automatic detection system for surface pitting of plastic masterbatch; Figure 5 A schematic diagram of the mounting bracket for an automatic detection system for surface pitting of plastic masterbatch; Figure 6 A schematic diagram of the structure of a fixed crossbar in an automatic detection system for surface pitting of plastic masterbatch; Figure 7 A schematic diagram of the fixing buckle structure of an automatic detection system for surface pitting of plastic masterbatch; Figure 8 A schematic diagram of the structure of a limiting plate in an automatic detection system for surface pitting of plastic masterbatch; Figure 9 A system framework diagram of an automatic detection system for surface pitting in plastic masterbatch; Figure 10 This is a flowchart illustrating an automatic detection method for surface pitting in plastic masterbatch. In the attached diagram: 100, sheet forming device; 110, drive motor; 120, hopper; 130, material cylinder; 140, die head; 200, sheet conveying device; 210, servo motor; 220, conveyor belt; 300, vision inspection system; 310, image acquisition component; 311, lighting device; 312, imaging device; 320, image acquisition trigger component; 321, incremental encoder; 322, signal processing module; 330, mounting bracket; 331, first mounting slot; 332, second mounting slot; 333, crossbar; 334, upright; 340, fixed crossbar; 350, fixing buckle; 351, first connecting hole; 360, limiting piece; 361, first... 362. Second connecting part; 363. Third connecting part; 364. Second connecting hole; 400. Data processing system; 410. Image preprocessing submodule; 411. Region of interest extraction module; 412. Noise suppression module; 420. Feature extraction and segmentation submodule; 430. Defect judgment submodule; 500. Air cooling device; 600. Software system; 610. Algorithm module; 620. Application module; 621. Image acquisition control submodule; 622. Data management submodule; 623. Human-computer interaction submodule; 630. Presentation layer; 640. Business logic layer; 650. Device interaction layer; 660. Image processing layer; 700. Cleaning and protection structure. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0023] The present invention will be further described below with reference to specific embodiments. The accompanying drawings are for illustrative purposes only, representing schematic diagrams rather than actual physical objects, and should not be construed as limiting the scope of this patent. To better illustrate the embodiments of the present invention, some components in the drawings may be omitted, enlarged, or reduced, and do not represent the actual dimensions of the product. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0024] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present patent. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0025] Example 1 This embodiment is the first embodiment of an automatic detection system for surface pitting in plastic masterbatch, including a sheet forming device 100, a sheet conveying device 200, a vision inspection system 300, and a data processing system 400. Figures 1 to 3 As shown in the figure. The automatic detection system for surface pitting on plastic masterbatch in this embodiment can be directly integrated into the plastic masterbatch production line. The plastic masterbatch produced by the production line is quantitatively and online fed into the sheet forming device 100. The sheet forming device 100 forms the plastic masterbatch into sheets according to a set temperature and pressure. The sheets are continuously conveyed to the vision inspection system 300 by the sheet conveying device 200 at a set interval and speed. The vision inspection system 300 performs visual inspection on each sheet. The images detected by the vision inspection system 300 are analyzed and processed by the data processing system 400 to obtain the characteristic parameters and spatial location of the pitting defect areas. In this embodiment, the plastic masterbatch can specifically be thermoplastic plastic masterbatch such as PVC masterbatch, PE masterbatch, or PP masterbatch.

[0026] The sheet forming device 100 is a screw extruder with a slit-shaped die, used to extrude plastic masterbatch into sheets of the target size. The sheet forming device 100 is located at one end of the sheet conveying device 200. Specifically, in this embodiment, the sheet forming device 100 includes a drive motor 110, a screw, a hopper 120, a barrel 130, and a die head 140. The drive motor 110 is fixedly installed, and the screw is connected to the drive motor 110 and located inside the barrel 130. The hopper 120 is installed above the barrel 130, and the die head 140 is installed at the end of the barrel 130 and has a slit-shaped die. In this embodiment, the flow channel in the die head 140 can be mirror-finished to ensure forming accuracy, and the die head 140 can be replaced according to the shape and specifications of the sheet to be formed. In this embodiment, the plastic masterbatch enters the barrel 130 from the hopper 120. Under the rotation and pushing of the screw and the heating of the barrel 130, it melts into a melt and is extruded through the die head 140 to obtain a sheet of the target size. The sheet falls onto the sheet conveying device 200 under the action of gravity, preparing for the subsequent sheet conveying and inspection.

[0027] The sheet conveying device 200 includes a servo motor 210 with a built-in encoder and a conveyor belt 220 driven by the servo motor 210 at a set speed. The conveyor belt 220 receives the sheets extruded by the sheet forming device 100 and conveys multiple sheets to the vision inspection system 300 at a set interval and speed. The servo motor 210 has a built-in encoder that monitors the speed in real time. The servo motor 210 adjusts its operating state in real time according to the speed monitored by the encoder, thereby achieving closed-loop feedback control. This ensures that the sheets are conveyed at a uniform speed and stably on the conveyor belt 220, avoiding imaging deviations caused by speed fluctuations or vibrations, thus helping the vision inspection system 300 achieve high-quality imaging. Furthermore, the pulse signal output by the encoder in real time can be used to calculate the sheet conveying speed and convert it into a scanning signal for the vision inspection system 300. The vision inspection system 300 acquires images using an equal-interval scanning method, further contributing to high-quality imaging.

[0028] To cool the sheet surface and stabilize it below 100°C, thus preventing the high surface temperature from affecting detection accuracy, this embodiment includes several air-cooling devices 500 evenly distributed above the sheet conveying device 200. Specifically, each air-cooling device 500 is a fan array arranged at uniform intervals, positioned above the conveyor belt 220's transport path. To create a relatively enclosed detection space and reduce interference from external light and dust, this embodiment also includes a cleaning and protective structure 700, with at least the sheet conveying device 200 and the vision inspection system 300 located inside the cleaning and protective structure 700. The cleaning and protective structure 700 can be a protective curtain made of transparent anti-static material, sliding along a rail and opening / closing as needed; alternatively, the detection system can be placed on a detection platform with a sliding groove, the rail engaging with the groove for sliding opening and closing. However, it should be noted that the air-cooling devices 500 and the cleaning and protective structure 700 in this embodiment are preferred options to further improve detection accuracy and are not intended to limit the invention.

[0029] The vision inspection system 300 is located at the other end of the sheet transport device 200. It includes an image acquisition component 310 and an image acquisition trigger component 320 connected by signals. The image acquisition trigger component 320 is connected to an encoder signal and sends a signal to the image acquisition component 310 to perform line scanning. The image acquisition component 310 is located above the conveyor belt 220 and is used to acquire sheet images. During line scanning, the direction of the line is perpendicular to the transport direction of the sheet. Figure 3 As shown.

[0030] In this embodiment, the image acquisition triggering component 320 includes an incremental encoder 321 and a signal processing module 322. The encoder of the servo motor 210 outputs pulse signals to the incremental encoder 321 in real time. The incremental encoder 321 receives the pulse signals and transmits them to the signal processing module 322. The signal processing module 322 calculates the transmission speed of the sheet based on the pulse signals and converts the transmission speed signal of the sheet into a scanning signal for the image acquisition component 310. The image acquisition component 310 acquires images in an equal-interval scanning manner, which helps the vision inspection system 300 achieve high-quality imaging.

[0031] In addition, in this embodiment, the image acquisition component 310 includes an illumination device 311 and an imaging device 312 located on both sides of a vertical line: the central axis of the illumination device 311 is provided at an angle with the vertical line. The imaging device 312 has an angle between its central axis and the vertical line. ;in, , In this embodiment, the illumination device 311 uses a coaxial light source, and the light emitted by the coaxial light source can uniformly cover the surface of the sheet. The imaging device 312 consists of an 8K resolution high-speed linear array camera and a telephoto macro lens. The focal length of the lens is set according to the width of the sheet and the detection accuracy. The lens is also equipped with an aperture adjustment ring and an exposure control module. The aperture adjustment ring can manually adjust the aperture size, and the exposure control module automatically adjusts the exposure time and camera gain according to the sheet transmission speed and ambient light intensity to ensure that the acquired image is clear and free of motion blur, and to ensure that the imaging accuracy reaches 0.06mm / pixel. The illumination device 311 and the imaging device 312 form a reflective structure, so that the pitting defects present obvious grayscale differences in the image.

[0032] The data processing system 400 receives sheet images and analyzes them to obtain the feature parameters and spatial location of the pitting defect area. The feature parameters include area, perimeter, and roundness. Specifically, in this embodiment, the data processing system 400 includes an image preprocessing submodule 410, a feature extraction and segmentation submodule 420, and a defect determination submodule 430, used to process, analyze, and manage the images acquired by the vision inspection system 300 to identify and determine pitting defects. The data processing system 400 in this embodiment can specifically run on an industrial control computer.

[0033] The image preprocessing submodule 410 includes a region of interest extraction module 411 and a noise suppression module 412. The region of interest extraction module 411 processes the sheet image and extracts a first image, while the noise suppression module 412 filters noise from the first image to obtain a second image that retains the details of the pitted edges. Specifically, the region of interest extraction module 411 determines the edges of the sheet using an adaptive threshold segmentation algorithm and removes background areas using Blob analysis to extract the first image. The noise suppression module 412 uses a nonlocal mean filtering algorithm to filter salt-and-pepper noise and Gaussian noise from the region of interest image, retaining the details of the pitted edges to obtain the second image. The feature extraction and segmentation submodule 420 is used to calculate the local standard deviation and gradient magnitude of the second image to construct a texture feature map. It uses a region growing algorithm to binarize the map, separating the pitted defects from the background. It is also used to remove isolated noise points through morphological opening and closing operations, connect broken pitted defect areas, and restore the true shape of the pits. Defect Judgment Submodule 430: Used to extract feature parameters of each pit defect region using connected component analysis technology, and compare the feature parameters with a preset judgment threshold to identify and classify the pit defect region.

[0034] The aforementioned data processing system 400 is an algorithm module 610 running on an industrial control computer. The industrial control computer also runs an application module 620. The application module 620 includes an image acquisition and control submodule 621, a data management submodule 622, and a human-computer interaction submodule 623, such as... Figure 4 As shown. Specifically, in this embodiment, the image acquisition control submodule 621: develops a driver program based on MVSSDK, connects to the imaging device 312 through a software interface, and realizes functions such as camera parameter configuration and real-time image transmission, ensuring that the acquisition frame rate is synchronized with the sheet material transmission speed; the data management submodule 622: the data storage module uses a database to store defect information, including the spatial coordinates, feature parameters, and detection time of the pit defect area, as well as sheet material images named and saved according to time nodes, and sets an automatic cleaning mechanism to periodically delete images that have exceeded the storage period to release storage space. The human-computer interaction submodule 623: develops an operation interface based on the WPF framework. The interface includes a real-time detection screen display area, a defect statistics report area, a parameter setting area, and a historical data query area. Users can adjust the detection parameters through the interface. When the number of pit defects within the cumulative set length exceeds the set threshold, the alarm module will issue an alarm prompt, and the interface will automatically alarm and highlight the abnormal area for convenient timely handling by operators.

[0035] like Figure 4As shown, in this embodiment, the sheet forming device 100, sheet conveying device 200, air cooling device 500, and cleaning and protection structure 700 constitute a mechanical system. The vision inspection system 300 serves as the vision system, and the algorithm module 610 and application module 620 serve as the software system 600. The mechanical system, vision system, and software system 600 work together. The mechanical system ensures stable sheet conveying and a clean inspection environment, the vision system acquires high-quality images, and the software system 600 performs precise analysis and management of the images. Together, they achieve real-time, high-precision detection of surface pitting on the sheet, meeting the quality control requirements of the production line.

[0036] Example 2 This embodiment is the second embodiment of the automatic detection system for surface pitting of plastic masterbatch. This embodiment is similar to the first embodiment, except that: the automatic detection system for surface pitting of plastic masterbatch in this embodiment further includes a mounting bracket 330, a first fixed frame, and a second fixed frame; the mounting bracket 330 has a first mounting groove 331 and a second mounting groove 332, the second mounting groove 332 being higher than the first mounting groove 331; an illumination device 311 is slidably mounted in the first mounting groove 331 via the first fixed frame; an imaging device 312 is slidably mounted in the second mounting groove 332 via the second fixed frame, the sliding direction of the illumination device 311 is perpendicular to the sliding direction of the imaging device 312, and the illumination device 311 and the imaging device 312 are located above the conveyor belt 220, as shown below. Figure 5 As shown. The sliding direction of the lighting device 311 is horizontal, which can adjust the horizontal position of the lighting device 311. The sliding direction of the imaging device 312 is vertical, which can adjust the vertical height of the imaging device 312.

[0037] Specifically, the mounting bracket 330 in this embodiment includes several horizontal bars 333 and several vertical bars 334, which form two cubic structures. The space inside the bottom cubic structure allows the sheet conveying device 200 to pass through, and the top cubic structure allows the lighting device 311 and the imaging device 312 to be positioned above the sheet to be inspected. The first mounting groove 331 is located on two horizontal bars 333 opposite each other on the top surface of the top cubic structure. Two vertical bars 334 opposite each other on one side adjacent to the top surface extend upward to form columns with a height higher than the top surface of the cubic structure. The second mounting groove 332 is located on the two opposite columns. The central axes of the first mounting groove 331 and the second mounting groove 332 are perpendicular. The length of the second mounting groove 332 can be greater than the length of the first mounting groove 331. The height adjustment stroke of the imaging device 312 is greater than the horizontal position adjustment stroke of the lighting device 311. Figure 5As shown. Both the first mounting groove 331 and the second mounting groove 332 are oblong grooves. The first fixed frame and the second fixed frame slide within the first mounting groove 331 and the second mounting groove 332, respectively. A first locking member and a second locking member can be respectively provided. When the first fixed frame slides to the desired position, the position of the lighting device 311 is locked by the first locking member. When the second fixed frame slides to the desired position, the position of the imaging device 312 is locked by the second locking member. The positions of the lighting device 311 and the imaging device 312 are adjustable, which can adapt to imaging of sheets of different thicknesses and has a wide range of applications.

[0038] like Figures 6 to 8 As shown, both the first and second fixed frames include a fixed crossbar 340, a fixed buckle 350, and a limiting piece 360. The end of the fixed crossbar 340 is slidably installed in the first mounting groove 331 or the second mounting groove 332. The limiting piece 360 ​​includes a first connecting part 361 fixedly connected to the fixed crossbar 340, a second connecting part 362 fixedly connected to the lighting device 311 or the imaging device 312, and a third connecting part 363 disposed between the first connecting part 361 and the second connecting part 362. An included angle θ is formed between the second connecting part 362 and the third connecting part 363. The included angle θ is specifically the included angle between the center lines of the second connecting part 362 and the third connecting part 363; the fixing buckle 350 is fixedly connected to the second connecting part 362, and the fixing buckle 350 is fixedly connected to the lighting device 311 or the imaging device 312.

[0039] Specifically, in this embodiment, the fixing buckle 350 is a U-shaped frame, with two first connecting holes 351 on each of the two opposite sides of the U-shaped frame. Two second connecting holes 364 are also provided on the second connecting portion 362 of the limiting piece 360. During installation, the first connecting holes 351 and the second connecting holes 364 are opposite each other and fixed to the side of the lighting device 311 or the imaging device 312. The fixing crossbar 340 has threaded sections at both ends. The first locking member and the second locking member are both locking nuts that mate with the threaded sections. The threaded sections pass through the first mounting groove 331 or the second mounting groove 332 and are locked by the first locking member or the second locking member. The first connecting part 361 is welded to the fixed crossbar 340 or fixedly connected to the fixed crossbar 340 in other ways. The fixing buckle 350 is fitted onto the outer periphery of the lighting device 311 or the imaging device 312. The limiting piece 360, the fixing buckle 350, and the lighting device 311 or the imaging device 312 are fixed together by the connecting piece passing through the first connecting hole 351 and the second connecting hole 364, forming a stable support system. When it is necessary to change the angle of the lighting device 311 or the imaging device 312, this can be achieved by replacing the limiting piece 360 ​​with one of different shapes.

[0040] Example 3 This embodiment is the third embodiment of the automatic detection system for surface pitting of plastic masterbatch. This embodiment is similar to Embodiment 1, except that the software system 600 in this embodiment has a hierarchical architecture design, and each layer collaborates to achieve online detection and linkage control of sheet pitting defects, covering the presentation layer 630, the business logic layer 640, the device interaction layer 650, and the image processing layer 660. Each layer uses C# as the core interaction language to form a closed-loop control system, as Figure 9 shown.

[0041] Presentation layer 630: Builds an interface based on the WPF framework, defines interface elements through the MainWindow class and XAML layout. The specific presentation interface is as follows: The original image is displayed on the left side of the interface, and the processed image with superimposed defect marks is displayed on the right side of the interface, realizing two-way data binding to update the captured image, processing results, pitting quantity, historical data, etc.; It may also include a detection data trend display interface, integrating the LiveCharts component to draw a scatter plot, mapping the sheet length on the horizontal axis and the pitting quantity on the vertical axis to display the detection data trend; It may also include a setting interface, where the setting interface can set values such as the plastic model to be detected, detection length, detection threshold, pitting size, etc., and select the corresponding image processing algorithm parameters configuration for different grades of plastics.

[0042] Business logic layer 640: Coordinates data flow and instruction processing among layers, realizes defect determination, data management, and process scheduling, and at the same time supports calling corresponding image processing algorithm parameters according to different grades of plastics and pitting types; Specifically, the implementation method of defect determination is as follows: When the quantity of various types of pitting within a certain cumulative length is compared with the preset threshold corresponding to this grade of plastic respectively, an alarm is triggered when it exceeds the preset threshold; The implementation method of data management is as follows: Records and saves detection data, and the detection data includes timestamp, grade, pitting type, quantity, length, and corresponding algorithm parameters, automatically deletes image files stored for more than a certain period to optimize the storage space; The implementation method of process scheduling is as follows: Realizes a closed-loop control of "image acquisition → preprocessing → defect detection → result recording → alarm judgment" through thread control, dynamically adjusts the processing rhythm through the timestamp difference, and can switch the corresponding processing logic according to the plastic grade and pitting type.

[0043] Device interaction layer 650: Calls the interface of the Hikvision SDK of the image acquisition component 310 through C#, analyzes device information, performs acquisition configuration, and sets the acquisition frame rate, image buffer size, and controls the light source brightness according to the plastic characteristics of different grades; Establishes communication with the screw extruder through the ModbusTCP protocol. When an abnormality is detected, it sends an NG signal to the screw extruder, triggers an audible and visual alarm and a shutdown operation, and at the same time displays a fault shutdown interface on the presentation layer 630, waits for the signal to be cleared after the fault is processed, and resumes operation.

[0044] Image processing layer 660: Adapts different image processing algorithms for different grades of plastics. The image processing algorithms include image preprocessing methods, feature extraction and segmentation methods, and defect determination methods.

[0045] Example 4 This embodiment is an example of an automatic detection method for surface pitting on plastic masterbatch, implemented based on the automatic detection system for surface pitting on plastic masterbatch from any of Embodiments 1 to 3. Figure 10 As shown, the automatic detection method in this embodiment specifically includes the following steps: The steps of sheet forming are as follows: plastic masterbatch is added into the screw extruder and extruded by the screw extruder to obtain a sheet of the target size; The sheet conveying process is as follows: The sheet falls onto the conveyor belt 220, and the conveyor belt 220 transports multiple sheets to the vision inspection system 300 at a set spacing and speed. While the conveyor belt 220 is transporting the sheet, the surface of the sheet is cooled to below 100°C. At the same time, the speed of the servo motor 210 is monitored in real time by the encoder. If the speed of the servo motor 210 is >30 rpm, it is determined that the machine is being washed, and no processing is performed. Only the original image is displayed on the display layer 630. If the speed of the servo motor 210 is ≤30 rpm, the image is saved with the current time as the original image, and it is determined that the pitting defect detection is in progress. The steps of sheet imaging are as follows: The image acquisition trigger component 320 receives the signal from the encoder to trigger the image acquisition component 310 to perform line scanning on the sheet surface to obtain a sheet image; The data processing steps are as follows: Analyze and process the sheet image to obtain the characteristic parameters of the pitted defect area and the spatial location of the pitted defect area.

[0046] Specifically, the data processing steps include: The image preprocessing steps are as follows: The edges of the sheet are determined using an adaptive threshold segmentation algorithm, and background areas are removed using Blob analysis to extract the first image; a non-local mean filtering algorithm is used to filter salt-and-pepper noise and Gaussian noise in the region of interest image, preserving the details of the pitted edges to obtain the second image; more specifically, in this embodiment, a threshold segmentation of 200 combined with line fitting and area filtering is used to remove the 220 background of the conveyor belt to obtain the sheet area, and edge contraction areas are removed using average grayscale analysis; noise suppression combines 5*5 mean filtering and median filtering to eliminate noise and preserve edge information. The steps of feature extraction and segmentation are as follows: Calculate the local standard deviation and gradient magnitude of the second image to construct a texture feature map, set the corresponding segmentation threshold, and use the region growing algorithm to binarize the map to separate the pitting defects from the background; remove isolated noise points through morphological opening and closing operations, connect the broken pitting defect areas, and restore the true shape of the pitting; more specifically, in this embodiment, different feature extraction and segmentation methods are used for different types of pitting, such as circular, elongated, sharp / deep pit, and shallow pit / shallow protrusion, with circular and sharp / deep pit pits being given special attention.

[0047] The steps for defect identification are as follows: Connected component analysis is used to extract the feature parameters of each pit defect region, and the feature parameters are compared with the preset judgment threshold to identify and classify the pit defect regions.

[0048] In the specific implementation of the above embodiments, the technical features can be combined in any non-contradictory way. For the sake of brevity, not all possible combinations of the above technical features are described. However, as long as the combination of these technical features is not contradictory, it should be considered to be within the scope of this specification.

[0049] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art can make other variations or modifications based on the above description. It is neither necessary nor possible to exhaustively describe all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the claims of the present invention.

Claims

1. An automatic detection system for surface pitting on plastic masterbatch, characterized in that, Includes sheet forming equipment, sheet conveying equipment, vision inspection system, and data processing system: The sheet forming device is a screw extruder with a slit-shaped die, used to extrude plastic masterbatch into sheets of the target size. The sheet forming device is located at one end of the sheet conveying device. The sheet conveying device includes a servo motor with a built-in encoder and a conveyor belt driven by the servo motor at a set speed. The conveyor belt is used to receive the sheet extruded by the sheet forming device and to convey multiple sheets to the vision inspection system at a set spacing and a set speed. The visual inspection system is located at the other end of the sheet conveying device. It includes an image acquisition component and an image acquisition triggering component connected by a signal. The image acquisition triggering component is connected to the encoder by a signal. The image acquisition triggering component is used to send a signal to the image acquisition component to perform line scanning. The image acquisition component is located above the conveyor belt and is used to acquire sheet images. In the line scanning, the direction of the line is perpendicular to the conveying direction of the sheet. The data processing system is used to receive the sheet image and analyze it to obtain the feature parameters of the pitted defect area and the spatial location of the pitted defect area. The feature parameters include area, perimeter and roundness.

2. The automatic detection system for surface pitting of plastic masterbatch according to claim 1, characterized in that, The image acquisition component includes an illumination device and an imaging device located on both sides of a vertical line: The central axis of the lighting device forms an angle with the vertical line. ; The central axis of the imaging device forms an angle with the vertical line. ; in, , .

3. The automatic detection system for surface pitting of plastic masterbatch according to claim 2, characterized in that, It also includes a mounting bracket, a first fixing frame, and a second fixing frame: The mounting bracket is provided with a first mounting groove and a second mounting groove, wherein the second mounting groove is higher than the first mounting groove. The lighting device is slidably mounted in the first mounting slot via the first fixed frame; The imaging device is slidably mounted in the second mounting slot via the second fixed frame; Both the lighting device and the imaging device are located above the conveyor belt.

4. The automatic detection system for surface pitting of plastic masterbatch according to claim 3, characterized in that, Both the first and second fixed frames include a fixed crossbar, a fixed buckle, and a limiting piece. The end of the fixed crossbar is slidably installed in the first mounting groove or the second mounting groove; The limiting piece includes a first connecting part fixedly connected to a fixed crossbar, a second connecting part fixedly connected to an illumination device or an imaging device, and a third connecting part disposed between the first and second connecting parts, wherein an included angle θ is formed between the second and third connecting parts. ; The fixing buckle is fixedly connected to the second connecting part, and the fixing buckle is fixedly connected to the lighting device or imaging device.

5. The automatic detection system for surface pitting of plastic masterbatch according to claim 1, characterized in that, The data processing system includes an image preprocessing submodule, a feature extraction and segmentation submodule, and a defect determination submodule. The image preprocessing submodule includes a region of interest extraction module and a noise suppression module. The region of interest extraction module is used to process the sheet image and extract a first image. The noise suppression module is used to filter noise in the first image to obtain a second image that retains the details of the pitted edges. The feature extraction and segmentation submodule is used to process the second image and separate the pitted defect area from the background, and to connect the broken pitted defect areas to restore the true shape of the pits. The defect determination submodule is used to analyze the feature parameters of the pitted defect area and compare the feature parameters with a preset threshold to obtain the defect classification.

6. The automatic detection system for surface pitting of plastic masterbatch according to claim 5, characterized in that, It also includes a data storage module and an alarm module: Data storage module: used to record and store the spatial coordinates and characteristic parameters of the pitted defect area; Alarm module: Used to issue an alarm when the number of pockmarks within a set length exceeds a set threshold.

7. The automatic detection system for surface pitting of plastic masterbatch according to any one of claims 1 to 6, characterized in that, Several air-cooling devices are evenly distributed above the sheet conveying device.

8. The automatic detection system for surface pitting of plastic masterbatch according to any one of claims 1 to 6, characterized in that, It also includes a cleaning and protective structure, with at least the sheet conveying device and the vision inspection system located inside the cleaning and protective structure.

9. An automatic detection method for surface pitting on plastic masterbatch, implemented based on the automatic detection system for surface pitting on plastic masterbatch according to any one of claims 1 to 8, characterized in that, Includes the following steps: The steps of sheet forming are as follows: plastic masterbatch is added into the screw extruder and extruded by the screw extruder to obtain a sheet of the target size; The sheet conveying process involves the sheet falling onto a conveyor belt, and the conveyor belt transporting multiple sheets to the vision inspection system at a set interval and a set speed. The steps of sheet imaging are as follows: The image acquisition trigger component receives the signal from the encoder and triggers the image acquisition component to perform line scanning on the sheet surface to obtain a sheet image; The data processing steps are as follows: Analyze and process the sheet image to obtain the characteristic parameters of the pitted defect area and the spatial location of the pitted defect area.

10. The automatic detection method for surface pitting of plastic masterbatch according to claim 9, characterized in that, The data processing steps include: The image preprocessing steps are as follows: the edges of the sheet are determined by an adaptive threshold segmentation algorithm, and the background area is removed by combining Blob analysis to extract the first image; a nonlocal mean filtering algorithm is used to filter salt-and-pepper noise and Gaussian noise in the region of interest image, while preserving the details of the pitted edges to obtain the second image; The steps of feature extraction and segmentation are as follows: calculate the local standard deviation and gradient magnitude of the second image to construct a texture feature map; use the region growing algorithm to binarize the map and separate the pitting defects from the background; remove isolated noise points through morphological opening and closing operations, connect the broken pitting defect regions, and restore the true shape of the pitting. The steps for defect determination are as follows: Connected component analysis is used to extract the feature parameters of each pit defect region, and the feature parameters are compared with a preset determination threshold to identify and classify the pit defect regions.

Images