External surface defect recognition system and method for monitoring punching of plastic parts

Abstract

The present invention relates to the field of industrial automation detection technology, and discloses an external surface defect recognition system for monitoring punching of plastic parts, including a real-time monitoring module, a defect recognition module and a quality management module. The external surface defect recognition system for monitoring punching of plastic parts connects a high-definition camera and a laser scanner through a real-time monitoring module to classify and collect the appearance image and hole depression data of plastic parts products. The defect recognition module calculates and generates a product model Cpmx without manual intervention, and has strong automatic adaptation to various types of plastic parts recognition capabilities. The defect recognition module calculates and generates a punching data group Cksj to achieve batch monitoring. The defect recognition module calculates and generates a defect data group Qxsj to accurately analyze and evaluate cracks and scratches. The quality management module is set with three rounds to judge the quality level of plastic parts products and generate corresponding monitoring reports. The detection accuracy is high and the product quality is more stable.

Description

Technical Field

[0001] The present invention relates to the field of industrial automation detection technology, and in particular to an outer surface defect recognition system and method for monitoring punching of plastic parts. Background Art

[0002] The plastic part punching process refers to a processing method in which a punch is applied to a part or component made of plastic by applying pressure to make the punch pass through the plastic material to form a hole. This process is widely used in automobile manufacturing, electronic appliances, furniture accessories and other light industrial products. The steps of the plastic part punching process are to first design the punching mold according to the product requirements, select the appropriate plastic material, and do the preparation work before punching. Then place the plastic part on the punching machine to ensure its correct position, clamp it and start the punching machine, so that the punch passes through the plastic part with a certain pressure to form the required hole. Finally, after the punching is completed, clean the burrs and debris around the punching hole and conduct a preliminary quality inspection. In order to ensure the quality and efficiency of the punching, it is necessary to strictly control the design and manufacturing accuracy of the mold. Different plastic materials have different mechanical properties. Selecting the right material is crucial to ensure the quality of the punching. Parameters such as punching speed and pressure also need to be adjusted according to the specific material and aperture size. In the process of punching plastic parts, due to the influence of various factors, common process defects such as burrs, flash, aperture deviation, hole shape deviation, surface scratches and material cracks may occur. Excessive plastic burrs or flash directly affect the beauty and assembly quality of plastic parts. Aperture and hole shape deviations will cause the actual product to not meet the design standards required by customers. Improper punching operation or mold wear will leave scratches on the surface of plastic parts. Excessive punching pressure or insufficient toughness of the plastic material itself may cause material cracking. Therefore, in order to ensure the high quality of the plastic part punching process, it is extremely important to monitor and identify process defects. An effective monitoring system can detect defects in production in real time, so as to make timely adjustments and corrections to avoid the production of batches of defective products. With the development of modern industrial automation monitoring technology, especially the advancement of machine vision technology, monitoring and identification technology is becoming more and more intelligent, high-precision and automated.

[0003] At present, the traditional external surface defect recognition system used for plastic part punching monitoring needs to cope with flexible and changeable product types, and requires more manual intervention for system configuration. In actual application, the accuracy of the image captured by the detection device is easily affected by external factors such as ambient light and temperature. Tiny pit defects on the surface of plastic parts are not easy to be identified and discovered, resulting in uneven product quality. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In view of the shortcomings of the prior art, the present invention provides a system and method for identifying external surface defects for monitoring punching of plastic parts. The system and method have the advantages of automatic adaptation to various types of plastic parts, strong identification capability, high detection accuracy and more stable product quality. The system solves the problem that the traditional system for identifying external surface defects for monitoring punching of plastic parts requires manual intervention to adjust configuration parameters, and small defects are difficult to identify and have low detection accuracy.

[0006] (II) Technical solution

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] According to a first aspect of the present invention, a system for identifying external surface defects for monitoring punching of plastic parts is provided. The system comprises:

[0009] Real-time monitoring module, defect identification module and quality management module;

[0010] The real-time monitoring module includes an appearance monitoring unit and a punching monitoring unit; the appearance monitoring unit collects appearance data sets through a network connection with a high-definition camera and transmits them to the defect recognition module through the network; the appearance data sets include appearance images of plastic products at all angles, and the expression of the appearance data sets is: j represents the shooting angle of the HD camera, to The appearance images of each plastic product at different angles correspond in sequence; the punching monitoring unit collects the punching data set through a network connection with a laser scanner and transmits it to the defect recognition module through the network; the punching data set includes the laser scanning data of all the outer surfaces of the plastic product, and the expression of the punching data set is: m represents the total number of faces on the surface recognized by the laser scanner, s represents the total number of holes and depressions recognized by the laser scanner, to The hole depression data corresponding to different planes of each plastic product are sequentially connected to the defect recognition module via a network;

[0011] The defect recognition module includes an appearance analysis unit, a punching analysis unit and a defect analysis unit; the appearance analysis unit calculates and generates a product model Cpmx according to the appearance data set, and transmits it to the punching analysis unit through the network; the punching analysis unit calculates and generates a punching data group Cksj according to the product model Cpmx and the punching data set, and transmits it to the defect analysis unit through the network; the defect analysis unit calculates and generates a defect data group Qxsj according to the product model Cpmx and the punching data set; the defect recognition module is connected to the quality management module through the network;

[0012] The quality management module determines the quality level of the plastic parts product according to the product model Cpmx, the punching data group Cksj and the defect data group Qxsj, and generates a corresponding monitoring report.

[0013] Furthermore, the appearance analysis unit calculates and generates a product model Cpmx according to the appearance data set, including:

[0014] According to the appearance data set, the appearance images of a single plastic product at different angles are counted, denoted as C, where Then, according to the appearance images C of a single plastic product at different angles, high-definition images at different angles are screened frame by frame, denoted as X, where X = {X1, X2, X3, ..., X n}, X1 to X n High-definition images of each angle of the appearance of a single plastic product are sequentially displayed;

[0015] Cpmx=∑ Python (X1+X2+X3+...+X n

[0016] In the formula, Cpmx represents the product model, Python represents the Python algorithm model, and OpenCV and PCL libraries are used to fuse two-dimensional images from different perspectives to construct a three-dimensional structure, X1+X2+X3+...+X n High-definition images showing the appearance of a single plastic product from all angles,∑ Python (X1+X2+X3+...+X n ) means that high-definition images of the appearance of a single plastic product from all angles are substituted into the Python algorithm model to construct the three-dimensional structure of a single plastic product, which is the product model.

[0017] Furthermore, the punching analysis unit calculates and generates a punching data set Cksj according to the product model Cpmx and the punching data set, including:

[0018] According to the product model Cpmx, the total number of holes identified by the HD camera on the appearance of a single plastic product is counted and recorded as D sx ;

[0019] According to the punching data set, the total number of holes and depressions identified by the laser scanner on the appearance of a single plastic product is counted and recorded as D jg ,in, Then, based on the hole data set, the shapes, diameters, and depths of all the holes in the appearance of the same-shaped plastic parts are counted, and recorded as D xzs , where D xzs ={XZS1, XZS2, XZS3,...,XZS t}, XZS1 to XZSn The shape, diameter and depth of the holes and depressions on the exterior of each plastic product with the same shape correspond in sequence;

[0020]

[0021] In the formula, Cksj represents the punching data group, D sx It indicates the total number of holes detected by a high-definition camera on the appearance of a single plastic product. jg It indicates the total number of holes and depressions detected by laser scanner on the appearance of a single plastic product. sx -D jg represents the difference between the total number of holes identified twice, t represents the total number of plastic products with the same shape, It indicates the mean shape, diameter and depth of holes in the plastic parts with the same shape.

[0022] Furthermore, the defect analysis unit calculates and generates a defect data set Qxsj according to the product model Cpmx and the punching data set, including:

[0023] The defect analysis unit is provided with a YOLOv5 target detection algorithm model, and uses the YOLOv5 target detection algorithm model to identify the high-definition images X of the appearance of a single plastic product at different angles and the hole depression data of each plastic product on different planes in the punching data set;

[0024]

[0025] In the formula, Qxsj represents the defect data set, YOLOv5 represents the YOLOv5 target detection algorithm model, and real-time monitoring of cracks and scratches on the appearance of plastic parts. It represents high-definition images of the appearance of a single plastic product at different angles and hole and depression data on different planes. The cracks and scratches on the appearance of plastic products monitored in real time by the YOLOv5 target detection algorithm model are the defect data set.

[0026] Further: the quality management module compares the product model Cpmx with the preset processing model. If the product model Cpmx features are different from the preset processing model features, the quality management module determines that the plastic product does not meet the processing design standards and generates a model defect report.

[0027] Furthermore: the quality management module compares the punching data group Cksj with a preset value. If the punching data group Cksj is greater than the preset value, the quality management module determines that there are errors in the number, shape, diameter and depth of punching of the plastic product, and generates a hole defect report.

[0028] Further: when there is data in the defect data group Qxsj, the quality management module determines that there are cracks and scratches on the appearance of the plastic product and generates a quality defect report.

[0029] According to a second aspect of the present invention, a method for identifying external surface defects for monitoring punching of plastic parts is provided. The method comprises:

[0030] The plastic products to be identified are sequentially conveyed to a closed monitoring black box via a conveyor belt. A high-definition camera and a laser scanner are installed in the monitoring black box. The real-time monitoring module connects the high-definition camera and the laser scanner through a network to classify and collect the appearance images of the plastic products at all angles and the hole and depression data of the plastic products on different planes.

[0031] The defect recognition module constructs a three-dimensional product model Cpmx based on the appearance images of the plastic product from all angles. The quality management module compares the product model Cpmx with the preset processing model. If the features of the product model Cpmx are different from the features of the preset processing model, the recognition step is stopped and a model defect report is generated. If the features of the product model Cpmx are the same as the features of the preset processing model, the defect recognition module continues the recognition step.

[0032] The defect recognition module analyzes the number, shape, diameter and depth of the punching holes on the appearance of the plastic product according to the product model Cpmx and the hole depression data on different planes of the plastic product, and calculates and generates the punching data group Cksj. The quality management module compares the punching data group Cksj with the preset value. If the punching data group Cksj is greater than the preset value, the recognition step is stopped and a hole defect report is generated. If the punching data group Cksj is less than or equal to the preset value, the defect recognition module continues the recognition step.

[0033] The defect recognition module is equipped with a YOLOv5 target detection algorithm model. It combines the product model Cpmx and the hole and depression data of different planes of the plastic product to analyze the cracks and scratches on the appearance of the plastic product, and calculates and generates a defect data group Qxsj. If there is data in the defect data group Qxsj, the quality management module generates a quality defect report. If there is no data in the defect data group Qxsj, there is no defect in the appearance of the plastic product.

[0034] According to a third aspect of the present invention, an electronic device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the outer surface defect recognition method for monitoring punching of plastic parts is executed.

[0035] According to a fourth aspect of the present invention, a storage medium is provided, which stores a computer program that can be executed by one or more processors and can be used to implement the external surface defect recognition method for monitoring punching of plastic parts.

[0036] Compared with the prior art, the present invention provides a system and method for identifying external surface defects for monitoring punching of plastic parts, which has the following beneficial effects:

[0037] 1. The present invention sets an appearance monitoring unit and a punching monitoring unit through a real-time monitoring module, and conveys the plastic products to be identified to a closed monitoring black box in turn through a conveyor belt. A high-definition camera and a laser scanner are arranged in the monitoring black box. The real-time monitoring module connects the high-definition camera and the laser scanner through a network to classify and collect appearance images of plastic products at all angles and hole depression data of plastic products on different planes, thereby greatly improving the subsequent automatic identification efficiency. The defect identification module includes an appearance analysis unit, a punching analysis unit and a defect analysis unit. The appearance analysis unit calculates and generates a product model Cpmx according to the appearance data set. There is no need for manual intervention to adjust the parameter configuration of the defect identification system. The automatic adaptation to various types of plastic parts is strong in recognition capability, and it can quickly determine whether there is a position error or omission problem in the product holes in the subsequent period.

[0038] 2. The present invention calculates and generates the punching data group Cksj according to the product model Cpmx and the punching data set through the punching analysis unit, so as to realize batch monitoring of plastic products of the same shape, which helps to quickly judge whether the number of punching holes is correct, whether there are burrs in the hole shape, whether the diameter is uniform, and whether the hole depth meets the processing standards, which is beneficial to improve the accuracy of product quality control. The defect analysis unit calculates and generates the defect data group Qxsj according to the product model Cpmx and the punching data set, so as to accurately analyze and evaluate the severity of cracks and scratches. The quality management module sets three rounds to judge the quality level of plastic products according to the product model Cpmx, the punching data group Cksj and the defect data group Qxsj, and generates corresponding monitoring reports. The detection accuracy is high and the product quality is more stable. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The above and other features, advantages and aspects of the embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are used to better understand the present invention and do not constitute a limitation of the present invention. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, wherein:

[0040] Figure 1 A schematic structural diagram of an outer surface defect recognition system for monitoring punching of a plastic part according to an embodiment of the present invention is shown;

[0041] Figure 2 A schematic flow chart of an outer surface defect recognition method for monitoring punching of a plastic part according to an embodiment of the present invention is shown;

[0042] Figure 3A block diagram of an exemplary electronic device capable of implementing embodiments of the present invention is shown. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] Figure 1 A schematic diagram of the structure of an outer surface defect recognition system for monitoring punching of a plastic part according to an embodiment of the present invention is shown, and the system includes:

[0045] Real-time monitoring module 110, defect identification module 120 and quality management module 130;

[0046] The real-time monitoring module 110 includes an appearance monitoring unit 111 and a punching monitoring unit 112. The appearance monitoring unit 111 collects appearance data sets through a network connection with a high-definition camera and transmits the data sets to the defect recognition module 120 through the network. The appearance data sets include appearance images of the plastic product from all angles. The expression of the appearance data sets is: j represents the shooting angle of the high-definition camera. There are multiple high-definition cameras in the monitoring black box. The specific shooting angles include the front, left, right, back, top and bottom, which can flexibly cope with various types of plastic products. to The appearance images of each plastic product at different angles are selected frame by frame based on the high-definition video, which retains the appearance details of the plastic product to the greatest extent, making subsequent analysis faster and reducing recognition errors.

[0047] The punching monitoring unit 112 collects the punching data set through a network connection with a laser scanner and transmits it to the defect recognition module 120 through the network. The punching data set includes the laser scanning data of all the outer surfaces of the plastic product. The expression of the punching data set is: m represents the total number of surfaces recognized by the laser scanner. The laser scanner scans the surface of the plastic product with a high-intensity laser beam and then captures the scan data through a high-precision sensor. It is more adaptable to plastic products with complex shapes and curved surfaces. s represents the total number of holes and depressions recognized by the laser scanner. to Corresponding to the hole depression data of different planes of each plastic product, the laser scanner can simultaneously obtain information such as the number, shape, diameter, depth and color of the holes on the surface of the target product from multiple angles, providing more comprehensive monitoring data for subsequent defect analysis;

[0048] The real-time monitoring module 110 is connected to the defect recognition module 120 via a network, and clearly classifies the appearance images of the plastic product at all angles and the hole and depression data of the plastic product at different planes, which greatly improves the subsequent automatic recognition efficiency;

[0049] The defect recognition module 120 includes an appearance analysis unit 121, a punching analysis unit 122 and a defect analysis unit 123. The appearance analysis unit 121 calculates and generates a product model Cpmx according to the appearance data set, and transmits it to the punching analysis unit 122 through the network. The calculation process is as follows:

[0050] According to the appearance data set, the appearance images of a single plastic product at different angles are counted, denoted as C, where Then, according to the appearance images C of a single plastic product at different angles, high-definition images at different angles are screened frame by frame, denoted as X, where X = {X1, X2, X3, ..., X n}, X1 to X n High-definition images of each angle of the appearance of a single plastic product are sequentially displayed;

[0051] Cpmx=∑ Python (X1+X2+X3+...+X n

[0052] In the formula, Cpmx represents the product model, Python represents the Python algorithm model, and OpenCV and PCL libraries are used to fuse two-dimensional images from different perspectives to construct a three-dimensional structure, X1+X2+X3+...+X n High-definition images showing the appearance of a single plastic product from all angles,∑ Python (X1+X2+X3+...+X n ) means that high-definition images of the appearance of a single plastic part from all angles are substituted into the Python algorithm model to construct a three-dimensional structure of a single plastic part, which is the product model. According to the product model Cpmx, the parameter configuration of the defect recognition system can be adjusted without manual intervention. It has strong automatic adaptation to various types of plastic parts recognition capabilities, and can quickly determine whether there is a position error or omission problem in the product holes.

[0053] The punching analysis unit 122 calculates and generates a punching data set Cksj according to the product model Cpmx and the punching data set, and transmits it to the defect analysis unit 123 through the network. The calculation process is as follows:

[0054] According to the product model Cpmx, the total number of holes identified by the HD camera on the appearance of a single plastic product is counted and recorded as D sx ;

[0055] According to the punching data set, the total number of holes and depressions identified by the laser scanner on the appearance of a single plastic product is counted and recorded as D jg ,in, Then, based on the hole data set, the shapes, diameters, and depths of all the holes in the appearance of the same-shaped plastic parts are counted, and recorded as D xzs , where D xzs ={XZS1, XZS2, XZS3,...,XZS t}, XZS1 to XZS n The shape, diameter and depth of the holes and depressions on the exterior of each plastic product with the same shape correspond in sequence;

[0056]

[0057] In the formula, Cksj represents the punching data group, D sx It indicates the total number of holes detected by a high-definition camera on the appearance of a single plastic product. jg It indicates the total number of holes and depressions detected by laser scanner on the appearance of a single plastic product. sx -D jg represents the difference between the total number of holes identified twice, t represents the total number of plastic products with the same shape, Indicates the average shape, diameter and depth of the holes in the appearance of plastic parts with the same shape. Based on the punching data group Cksj, batch monitoring of plastic parts with the same shape can be achieved, which helps to quickly determine whether the number of punching holes is correct, whether there are burrs in the hole shape, whether the diameter is uniform, and whether the hole depth meets the processing standards. It is beneficial to improve the accuracy of product quality control. High detection accuracy means more stable product quality.

[0058] The defect analysis unit 123 calculates and generates a defect data set Qxsj according to the product model Cpmx and the punching data set. The calculation process is as follows:

[0059] The defect analysis unit 123 is provided with a YOLOv5 target detection algorithm model, and uses the YOLOv5 target detection algorithm model to identify the high-definition image X of the appearance of a single plastic product at different angles and the hole depression data of each plastic product on different planes in the punching data set;

[0060]

[0061] In the formula, Qxsj represents the defect data set, YOLOv5 represents the YOLOv5 target detection algorithm model, and real-time monitoring of cracks and scratches on the appearance of plastic parts. It represents high-definition images of the appearance of a single plastic product at different angles and hole and depression data on different planes. The cracks and scratches on the appearance of plastic parts monitored in real time by the YOLOv5 target detection algorithm model are the defect data set. According to the defect data set Qxsj, the severity of the cracks and scratches can be accurately analyzed and evaluated, which is conducive to improving the accuracy of product quality control. The higher the detection accuracy, the more stable the product quality.

[0062] The defect identification module 120 is connected to the quality management module 130 via a network;

[0063] The quality management module 130 determines the quality level of the plastic part product according to the product model Cpmx, the punching data group Cksj and the defect data group Qxsj. The quality management module 130 compares the product model Cpmx with the preset processing model. This comparison is the first round of judgment. If the product model Cpmx feature is different from the preset processing model feature, the preset processing model is specifically a plastic part production design model. The difference between the two model features, for example, the actual plastic part product appearance punching position is wrong, the quality management module 130 determines that the plastic part product does not meet the processing design standard, generates a model defect report, and the quality The quantity management module 130 compares the punching data group Cksj with the preset value. This comparison is the second round of judgment. If the punching data group Cksj is greater than the preset value, the preset value is specifically the micron-level tolerance threshold set by the manager according to actual production needs. The quality management module 130 determines that there are errors in the number, shape, diameter and depth of punching of the plastic product, and generates a hole defect report. When there is data in the defect data group Qxsj, this comparison is the final round of judgment. The quality management module 130 determines that there are cracks and scratches on the appearance of the plastic product, and generates a quality defect report. The detection accuracy is high and the product quality is more stable.

[0064] Figure 2 A schematic diagram of the structure of a method for identifying outer surface defects for monitoring punching of a plastic part according to an embodiment of the present invention is shown, the method comprising:

[0065] S201, the plastic products to be identified are sequentially conveyed to a closed monitoring black box via a conveyor belt. A high-definition camera and a laser scanner are provided in the monitoring black box. The real-time monitoring module is connected to the high-definition camera and the laser scanner through a network to classify and collect appearance images of the plastic products at all angles and hole and depression data of the plastic products on different planes, thereby greatly improving the efficiency of subsequent automatic identification;

[0066] S202, the defect recognition module constructs a three-dimensional product model Cpmx according to the appearance images of the plastic product from all angles, and the quality management module compares the product model Cpmx with the preset processing model. If the product model Cpmx feature is different from the preset processing model feature, the recognition step is stopped and a model defect report is generated to quickly eliminate unqualified plastic products. The defect recognition efficiency is higher. If the product model Cpmx feature is the same as the preset processing model feature, the defect recognition module continues the recognition step, indicating that the target plastic product has passed the first round of judgment, and the automatic adaptation to multiple types of plastic parts has strong recognition capabilities;

[0067] S203, the defect recognition module analyzes the number, shape, diameter and depth of the punching holes on the appearance of the plastic product according to the product model Cpmx and the hole depression data of different planes of the plastic product, calculates and generates the punching data group Cksj, and the quality management module compares the punching data group Cksj with the preset value. If the punching data group Cksj is greater than the preset value, the recognition step is stopped and a hole defect report is generated to quickly eliminate unqualified plastic products, and the defect recognition efficiency is higher. If the punching data group Cksj is less than or equal to the preset value, the defect recognition module continues the recognition step, indicating that the target plastic product has passed the second round of judgment.

[0068] S204, the defect recognition module is provided with a YOLOv5 target detection algorithm model, which combines the product model Cpmx and the hole and depression data of different planes of the plastic product to analyze the cracks and scratches on the appearance of the plastic product, and calculates and generates a defect data group Qxsj. If there is data in the defect data group Qxsj, the quality management module generates a quality defect report to quickly eliminate unqualified plastic products, and the defect recognition efficiency is higher. If there is no data in the defect data group Qxsj, it means that the target plastic product has passed three rounds of judgment, and there is no defect in the appearance of the plastic product. The detection accuracy is high and the product quality is more stable.

[0069] According to an embodiment of the present invention, the present invention also provides an electronic device and a readable storage medium.

[0070] Figure 3 A schematic block diagram of an electronic device that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0071] The electronic device includes a computing unit 301, which can perform various appropriate actions and processes according to a computer program stored in a ROM 302 or a computer program loaded from a storage unit 308 into a RAM 303. In the RAM 303, various programs and data required for the operation of the electronic device can also be stored. The computing unit 301, the ROM 302, and the RAM 303 are connected to each other via a bus 304. An I / O interface 305 is also connected to the bus 304.

[0072] A number of components in the electronic device are connected to the I / O interface 305, including: an input unit 306, such as a keyboard, a mouse, etc.; an output unit 307, such as various types of displays, speakers, etc.; a storage unit 308, such as a disk, an optical disk, etc.; and a communication unit 309, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 309 allows the electronic device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0073] The computing unit 301 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 301 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 301 performs the various methods and processes described above. For example, in some embodiments, the outer surface defect recognition method for monitoring punching of plastic parts may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 308. In some embodiments, part or all of the computer program may be loaded and / or installed on an electronic device via ROM 302 and / or a communication unit 309. When the computer program is loaded into RAM 303 and executed by the computing unit 301, one or more steps of the outer surface defect recognition method for monitoring punching of plastic parts described above may be performed. Alternatively, in other embodiments, the computing unit 301 may be configured to execute the outer surface defect recognition method for monitoring punching of plastic parts in any other appropriate manner (eg, by means of firmware).

[0074] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0075] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0076] In the context of the present invention, a readable storage medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A readable storage medium may be a machine-readable signal medium or a machine-readable storage medium. A readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. More specific examples of readable storage media may include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0077] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0078] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communications network). Examples of communications networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0079] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0080] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An external surface defect recognition system for monitoring punching of plastic parts, characterized in that: include: Real-time monitoring module, defect identification module and quality management module; The real-time monitoring module includes an appearance monitoring unit and a punching monitoring unit; the appearance monitoring unit collects appearance data sets through a network connection with a high-definition camera and transmits them to the defect recognition module through the network; the appearance data sets include appearance images of plastic products at all angles, and the expression of the appearance data sets is: , Indicates the shooting angle of the HD camera. to The appearance images of each plastic product at different angles correspond in sequence; the punching monitoring unit collects the punching data set through a network connection with a laser scanner and transmits it to the defect recognition module through the network; the punching data set includes the laser scanning data of all the outer surfaces of the plastic product, and the expression of the punching data set is: , Indicates the total number of faces on the exterior surface recognized by the laser scanner. Indicates the total number of holes and depressions identified by the laser scanner. to The hole depression data corresponding to different planes of each plastic product are sequentially connected to the defect recognition module via a network; The defect recognition module includes an appearance analysis unit, a punching analysis unit and a defect analysis unit; the appearance analysis unit calculates and generates a product model based on the appearance data set. and transmitted to the punching analysis unit via the network; Product Model include: According to the appearance data set, the appearance images of a single plastic product at different angles are counted and recorded as ,in, , and then based on the appearance images of a single plastic product at different angles Filter out high-definition images from different angles frame by frame, denoted as ,in, , to High-definition images of each angle of the appearance of a single plastic product are sequentially displayed; In the formula, represents the product model, Python represents the Python algorithm model, and OpenCV and PCL libraries are used to fuse two-dimensional images from different perspectives to construct a three-dimensional structure. High-definition images showing the appearance of a single plastic product from all angles. It means that high-definition images of the appearance of a single plastic product from all angles are substituted into the Python algorithm model to construct the three-dimensional structure of the single plastic product, which is the product model; The punching analysis unit is based on the product model and punching data set, calculate and generate punching data set and transmitted to the defect analysis unit via the network; Punching data set include: According to product model , count the total number of holes identified by a high-definition camera on the appearance of a single plastic product, and record it as ; According to the punching data set, the total number of holes and depressions identified by the laser scanner on the appearance of a single plastic product is counted and recorded as ,in, , and then according to the hole data set, the shapes, diameters and depths of all the holes in the appearance of the same shape plastic parts are counted and recorded as ,in, , to The shape, diameter and depth of the holes and depressions on the exterior of each plastic product with the same shape correspond in sequence; In the formula, Represents a punching data group, Indicates the total number of holes detected by a high-definition camera on the appearance of a single plastic product. Indicates the total number of holes and depressions detected by a laser scanner on the appearance of a single plastic product. It represents the difference between the total number of holes identified twice. Indicates the total number of plastic products of the same shape. Indicates the average shape, diameter and depth of the holes and depressions in the appearance of plastic parts with the same shape; The defect analysis unit is based on the product model and punching data set, calculate and generate defect data set ; Defect Data Group include: The defect analysis unit is equipped with a YOLOv5 target detection algorithm model, and uses the YOLOv5 target detection algorithm model to identify high-definition images of the appearance of a single plastic product at different angles. and the hole depression data of different planes of each plastic product in the punching data set; In the formula, represents a defect data group, express Target detection algorithm model, real-time monitoring of cracks and scratches on the appearance of plastic products, It represents high-definition images of the appearance of a single plastic product at different angles and hole and depression data on different planes. express The cracks and scratches on the appearance of plastic parts monitored in real time by the target detection algorithm model are the defect data set; The defect identification module is connected to the quality management module via a network; The quality management module is based on the product model , Punching data group and defect data set , judge the quality level of plastic parts products and generate corresponding monitoring reports; The quality management module converts the product model Compare with the preset processing model. If the product model When the features are different from the preset processing model features, the quality management module determines that the plastic product does not meet the processing design standards and generates a model defect report; The quality management module converts the punching data into Compared with the preset value, if the punching data group When it is greater than the preset value, the quality management module determines that there are errors in the number, shape, diameter and depth of punching holes in the plastic product and generates a hole defect report; The defect data set When there is data in the quality management module, the quality management module determines that there are cracks and scratches on the appearance of the plastic product and generates a quality defect report.

2. A recognition method for the outer surface defect recognition system for monitoring punching of plastic parts according to claim 1, characterized in that: include: The plastic products to be identified are sequentially conveyed to a closed monitoring black box via a conveyor belt. A high-definition camera and a laser scanner are installed in the monitoring black box. The real-time monitoring module connects the high-definition camera and the laser scanner through a network to classify and collect the appearance images of the plastic products at all angles and the hole and depression data of the plastic products on different planes. The defect recognition module builds a three-dimensional product model based on the appearance images of the plastic product from all angles. , the quality management module converts the product model Compare with the preset processing model. If the product model If the feature is different from the preset processing model feature, the recognition step is stopped and a model defect report is generated. When the feature is the same as the preset processing model feature, the defect recognition module continues the recognition step; Defect recognition module based on product model Hole depression data of different planes of plastic products, analyze the number, shape, diameter and depth of punching holes on the appearance of plastic products, and calculate and generate punching data sets , the quality management module will punch the data group Compared with the preset value, if the punching data group When the value is greater than the preset value, the recognition step is stopped and a hole defect report is generated. When it is less than or equal to the preset value, the defect recognition module continues the recognition step; The defect recognition module is equipped with the YOLOv5 target detection algorithm model, combined with the product model Hole and depression data on different planes of plastic parts, analyze cracks and scratches on the appearance of plastic parts, and calculate and generate defect data sets , if the defect data set When there is data in the quality management module, the quality defect report is generated. When there is no data in the data, there is no defect in the appearance of the plastic product.

3. An electronic device, characterized in that: The invention comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the identification method as claimed in claim 2 is executed.

4. A storage medium, characterized in that: The computer program stored in the storage medium can be executed by one or more processors and can be used to implement the identification method as claimed in claim 2.

Images