CN117169282B - Quality detection method and system of composite film for medicine packaging - Google Patents
Quality detection method and system of composite film for medicine packaging Download PDFInfo
- Publication number
- CN117169282B CN117169282B CN202311420506.3A CN202311420506A CN117169282B CN 117169282 B CN117169282 B CN 117169282B CN 202311420506 A CN202311420506 A CN 202311420506A CN 117169282 B CN117169282 B CN 117169282B
- Authority
- CN
- China
- Prior art keywords
- result
- abnormal
- value
- coordinates
- coordinate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 42
- 239000002131 composite material Substances 0.000 title claims abstract description 28
- 239000003814 drug Substances 0.000 title abstract description 18
- 238000004806 packaging method and process Methods 0.000 title abstract description 14
- 230000002159 abnormal effect Effects 0.000 claims abstract description 140
- 238000000034 method Methods 0.000 claims abstract description 42
- 238000004458 analytical method Methods 0.000 claims abstract description 22
- 238000012795 verification Methods 0.000 claims abstract description 22
- 230000000007 visual effect Effects 0.000 claims abstract description 21
- 230000007547 defect Effects 0.000 claims abstract description 20
- 238000005259 measurement Methods 0.000 claims abstract description 17
- 238000005070 sampling Methods 0.000 claims description 73
- 238000003331 infrared imaging Methods 0.000 claims description 35
- 238000011156 evaluation Methods 0.000 claims description 29
- 238000013507 mapping Methods 0.000 claims description 24
- 238000012937 correction Methods 0.000 claims description 23
- 238000012545 processing Methods 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 16
- 238000009826 distribution Methods 0.000 claims description 14
- 230000003044 adaptive effect Effects 0.000 claims description 10
- 238000009512 pharmaceutical packaging Methods 0.000 claims description 8
- 238000009825 accumulation Methods 0.000 claims description 7
- 230000005856 abnormality Effects 0.000 claims description 6
- 238000007689 inspection Methods 0.000 claims description 6
- 238000012423 maintenance Methods 0.000 claims description 6
- 230000001172 regenerating effect Effects 0.000 claims description 4
- 230000037303 wrinkles Effects 0.000 abstract description 6
- 230000005540 biological transmission Effects 0.000 abstract description 4
- 230000001105 regulatory effect Effects 0.000 abstract description 2
- 238000004422 calculation algorithm Methods 0.000 description 5
- 229940079593 drug Drugs 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000006978 adaptation Effects 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 230000004888 barrier function Effects 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000011179 visual inspection Methods 0.000 description 2
- 241001584785 Anavitrinella pampinaria Species 0.000 description 1
- 230000001133 acceleration Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000005022 packaging material Substances 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000007789 sealing Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The application provides a quality detection method and a quality detection system of a composite film for medicine packaging, and relates to the technical field of quality detection, wherein the method comprises the following steps: configuring the speed of a transmission belt, setting a length measurement sensor based on a transmission node, determining the visual field interval and acquisition delay of the thermal infrared imager, performing actual stable analysis according to a zero point resetting result, generating a tolerant drift coordinate, constructing unified image coordinate recognition to obtain an abnormal coordinate and an abnormal region, performing tolerant expansion, performing image acquisition according to an expansion result, performing quality verification, and generating a quality detection result. The method mainly solves the technical problems of high equipment cost, low accuracy and reliability. It is difficult to accurately detect some minor defects such as bubbles, dirt, wrinkles. The speed of the conveyor belt is regulated and controlled, so that the thermal infrared imager can accurately detect an abnormal region and generate a quality detection result.
Description
Technical Field
The invention relates to the technical field of quality detection, in particular to a quality detection method and a quality detection system of a composite film for medicine packaging.
Background
Quality detection of medicine packages is an important link for ensuring safety and effectiveness of medicines. However, the quality inspection of pharmaceutical packages at the present stage has some problems and disadvantages. For drug packaging material selection, some pharmaceutical enterprises may be overly concerned about cost and neglect the quality and safety of the material. Secondly, the sealability and barrier properties of the pharmaceutical packaging are important factors for ensuring that the pharmaceutical does not deteriorate. Meanwhile, some medicine packages have poor barrier properties and cannot effectively isolate air, water vapor and the like, so that the quality of medicines is affected. In addition, the print quality of labels and instructions on pharmaceutical packaging is also a matter of concern.
In the prior art, quality detection is performed through appearance detection, and orthoscopic visual inspection is performed on a composite film for medicine packaging at a bright place of natural light. The defects of perforation, foreign matters, peculiar smell, adhesion, separation between composite layers, obvious damage, bubbles, wrinkles, dirt and the like are not required. Meanwhile, the heat-sealing part of the composite bag should be flat and have no virtual seal.
The prior art has the technical problems of high equipment cost, low accuracy and reliability in detection. For some minor defects, such as bubbles, dirt, wrinkles, etc., visual inspection methods may cause erroneous or missed inspection. Thus, more sophisticated detection devices or image processing techniques may be required to improve the accuracy and reliability of the detection.
Disclosure of Invention
The method mainly solves the technical problems of high equipment cost, low accuracy and reliability. It is difficult to accurately detect some minor defects such as bubbles, dirt, wrinkles.
In view of the foregoing problems, embodiments of the present application provide a method and a system for detecting quality of a composite film for packaging medicines, and in a first aspect, embodiments of the present application provide a method for detecting quality of a composite film for packaging medicines, the method including: configuring a conveyor belt speed, setting a length measurement sensor at a conveying node, configuring the length measurement sensor based on a stability coefficient of the conveyor belt, determining a measurement correction length, determining a visual field interval and acquisition delay of the thermal infrared imager, setting an acquisition time node of the thermal infrared imager according to the conveyor belt speed, resetting a zero point of the acquisition time node based on the measurement correction length, carrying out actual stability analysis of the conveyor belt based on a zero point resetting result, generating a tolerance drift coordinate, establishing a unified image coordinate system of an image acquired by the thermal infrared imager, identifying and obtaining an abnormal coordinate and an abnormal region, carrying out tolerance expansion on the abnormal coordinate and the abnormal region through the tolerance drift coordinate, carrying out microscopic infrared imaging acquisition according to a tolerance expansion result, recording a microscopic infrared imaging acquisition result, carrying out quality verification based on the microscopic infrared imaging acquisition result, resetting the abnormal coordinate and the abnormal region according to a quality verification result, generating an abnormal identifier, recording a resetting result and an identifier result, and generating a quality detection result.
In a second aspect, embodiments of the present application provide a quality inspection system for a composite film for pharmaceutical packaging, the system comprising: the system comprises a metering correction length determining module, a tolerance drift coordinate generating module, an abnormal coordinate and abnormal region identifying module, a micro infrared imaging acquisition result recording module, a micro infrared imaging acquisition result expanding module and a micro infrared imaging quality identification and verification result, wherein the metering correction length determining module is used for configuring the speed of a conveyor belt, setting a length metering sensor at a conveying node, configuring the length metering sensor based on the stability coefficient of the conveyor belt, determining the metering correction length, a visual field interval and an acquisition delay determining module, the visual field interval and the acquisition delay determining module are used for determining the visual field interval and the acquisition delay of the infrared imager, setting an acquisition time node of the infrared imager according to the speed of the conveyor belt, resetting a zero point of the acquisition time node based on the metering correction length, the tolerance drift coordinate generating module is used for carrying out actual stable analysis of the conveyor belt based on a zero point resetting result, generating a tolerance drift coordinate, generating an abnormal coordinate and an abnormal region identifying module, the abnormal coordinate and the abnormal region identifying module is used for establishing a unified image coordinate system of an infrared imager acquisition image based on the stability coefficient of the conveyor belt, and identifying the abnormal region, the micro infrared imaging acquisition result recording module is used for expanding the abnormal coordinate and the abnormal region based on the tolerance drift coordinate, the micro infrared imaging acquisition result based on the micro infrared imaging quality identification module is used for carrying out tolerance expansion, the micro infrared imaging acquisition result and the micro infrared imaging quality identification module is used for carrying out the quality verification result, and the abnormal region is used for generating an abnormal quality identification.
One or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages:
the application provides a quality detection method and a quality detection system of a composite film for medicine packaging, and relates to the technical field of quality detection, wherein the method comprises the following steps: configuring the speed of a transmission belt, setting a length measurement sensor based on a transmission node, determining the visual field interval and acquisition delay of the thermal infrared imager, performing actual stable analysis according to a zero point resetting result, generating a tolerant drift coordinate, constructing unified image coordinate recognition to obtain an abnormal coordinate and an abnormal region, performing tolerant expansion, performing image acquisition according to an expansion result, performing quality verification, and generating a quality detection result.
The method mainly solves the technical problems of high equipment cost, low accuracy and reliability. It is difficult to accurately detect some minor defects such as bubbles, dirt, wrinkles. The speed of the conveyor belt is regulated and controlled, so that the thermal infrared imager can accurately detect an abnormal region and generate a quality detection result.
The foregoing description is merely an overview of the technical solutions of the present application, and may be implemented according to the content of the specification in order to make the technical means of the present application more clearly understood, and in order to make the above-mentioned and other objects, features and advantages of the present application more clearly understood, the following detailed description of the present application will be given.
Drawings
For a clearer description of the present disclosure or of the prior art, the drawings used in the description of the embodiments or of the prior art will be briefly described, it being obvious that the drawings in the description below are only exemplary and that other drawings may be obtained, without inventive effort, by a person skilled in the art, from the provided drawings.
Fig. 1 is a schematic flow chart of a quality detection method of a composite film for drug packaging according to an embodiment of the present application;
fig. 2 is a schematic flow chart of a method for determining the abnormal region and the abnormal coordinates in the quality detection method of the composite film for medicine packaging according to the embodiment of the present application;
FIG. 3 is a schematic flow chart of a method for determining a minimum sampling point distance constraint in a quality detection method of a composite film for drug packaging according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a quality detection system of a composite film for medicine packaging according to an embodiment of the present application.
Reference numerals illustrate: the system comprises a metering correction length determining module 10, a visual field interval and acquisition delay determining module 20, a tolerance drift coordinate generating module 30, an abnormal coordinate and abnormal region identifying module 40, a microscopic infrared imaging acquisition result recording module 50, an abnormal identification generating module 60 and a quality detection result generating module 70.
Detailed Description
The method mainly solves the technical problems of high equipment cost, low accuracy and reliability. It is difficult to accurately detect some minor defects such as bubbles, dirt, wrinkles.
For a better understanding of the foregoing technical solutions, the following detailed description will be given with reference to the accompanying drawings and specific embodiments of the present invention:
example 1
A method for detecting the quality of a composite film for packaging medicines as shown in fig. 1, the method comprising:
configuring the speed of a conveyor belt, setting a length measurement sensor at a conveying node, configuring the length measurement sensor based on the stability coefficient of the conveyor belt, and determining a measurement correction length;
specifically, the application requirements are determined: these requirements may include transport speed, item size, number of items to be handled, etc. These factors will determine the size, speed, and motor type of the conveyor belt. Selecting an appropriate conveyor belt speed: a suitable conveyor speed is selected. Generally, the faster the speed, the greater the number of items processed per minute. But also the design load of the conveyor belt and the maximum speed of the motor. Determining the type of length metering sensor: based on the application requirements, the type of length metering sensor required is determined. The length measuring sensor includes an ultrasonic sensor, a photoelectric sensor, a capacitance sensor, etc. The characteristic sensitivity of different types of sensors to a target object is different, and the sensor configuration is performed based on the stability coefficient of the conveyor belt: the stability factor of the conveyor belt has a great influence on its running speed and stability. If the conveyor belt is unstable, inaccurate sensor readings may result. Therefore, in providing a length metering sensor, the stability factor of the conveyor belt needs to be taken into account to ensure that the sensor is able to accurately meter the length of the article. Determining a metering correction length: after the conveyor belt speed and length gauge sensor type are determined, a gauge correction length needs to be determined. This is to ensure that the sensor will give accurate results when measuring the length of the article. In general, the optimal metrology correction length needs to be determined experimentally.
Determining a visual field interval and acquisition delay of the thermal infrared imager, setting an acquisition time node of the thermal infrared imager according to the speed of the conveyor belt, and resetting a zero point of the acquisition time node based on the metering correction length;
specifically, a field of view interval of the thermal infrared imager is determined: first, the field of view of the thermal infrared imager, i.e., the size of the smallest to largest object that can be clearly detected, needs to be considered. The field of view may be selected based on the application requirements, such as the size of the article to be detected, the distribution of the articles on the conveyor belt, etc. And (3) determining acquisition delay: acquisition delay refers to the time difference between the perception of the target object by the thermal infrared imager and the final acquisition of the image. The acquisition delay may be determined based on the conveyor speed and the target object size. For example, if the conveyor belt speed is 10 meters per minute and the target object length is 1 meter, the acquisition delay may be set to 1 second to ensure that the image is acquired in time as the target object passes through the thermal imager. Setting an acquisition time node: based on the conveyor belt speed and the target object size, the acquisition time node can be determined. For example, if the conveyor belt is running 10 meters per minute, the target object is 1 meter long, the acquisition delay is 1 second, then the acquisition time node may be set to run once per minute. Resetting the zero point of the acquisition time node: the zero point of the acquisition time node is reset based on the metering correction length to ensure that the length of the image acquired each time is consistent with the length of the actual article. For example, if the metrology correction length is 1 meter, then the start time for each image acquisition may be set at the time the first 1 meter length of item is present on the conveyor.
Carrying out actual stability analysis of the conveyor belt based on the zero point resetting result to generate a tolerant drift coordinate;
specifically, item data on a conveyor belt is collected: and acquiring data of the articles on the conveying belt by using a thermal infrared imager or other sensor equipment. Such data may include the size, shape, temperature, etc. of the article. Processing the data and extracting features: these collected data are processed and analyzed to extract characteristics related to conveyor belt stability. These characteristics may include the position of the article on the conveyor belt, speed, acceleration, etc. Modeling and predicting: based on these features and the zero-point reset results, a neural network model is built to predict the stability of the conveyor belt. The model is trained using machine learning algorithms, such as support vector machines, neural networks, and the like. Generating tolerant drift coordinates: and generating tolerant drift coordinates according to the prediction result. Tolerant drift coordinates refer to allowing some degree of movement or offset of the article on the conveyor to ensure stable operation of the conveyor. The tolerance drift can be realized by adjusting parameters such as speed, tension and the like of the conveyor belt.
Establishing a unified image coordinate system of an image acquired by the thermal infrared imager, and identifying and obtaining abnormal coordinates and an abnormal region;
specifically, an image coordinate system is established: first, a unified image coordinate system of an image acquired by the thermal infrared imager needs to be established. This coordinate system may have an origin at the upper left corner, a transverse direction being the X-axis and a longitudinal direction being the Y-axis. By such a coordinate system, the position of each pixel point in the image can be determined. Image preprocessing: the collected infrared image is preprocessed to reduce image noise and improve image quality. This may include contrast enhancement, filtering, smoothing, etc. Anomaly detection algorithm: an appropriate anomaly detection algorithm is employed, such as pixel value based, texture based, band based, etc., to identify an anomaly region in the image. These algorithms can detect abnormal regions from statistical distributions of pixel values, textures, etc. features. Determining abnormal coordinates: from the result of the anomaly detection algorithm, coordinates of the anomaly region can be determined. These coordinates can be expressed as positions in the image coordinate system. Labeling an abnormal region: the infrared image is marked with an abnormal region for subsequent analysis and processing. This can be achieved by drawing a rectangle, a circle, or the like on the image, and adding text or a mark, or the like.
Performing forgiving expansion on the abnormal coordinates and the abnormal region through the forgiving drift coordinates, executing microscopic infrared imaging acquisition according to the forgiving expansion result, and recording microscopic infrared imaging acquisition result;
specifically, the tolerant drift coordinate expansion: and carrying out tolerance expansion on the abnormal coordinates and the abnormal region according to the tolerance drift coordinates generated before. This may be accomplished by adding a range of pixels around the outlier coordinates, which may be determined from the tolerance drift coordinates. Microscopic infrared imaging acquisition: and carrying out microscopic infrared imaging acquisition based on the image expanded by the tolerant drift coordinates. This involves adjusting parameters of the thermal infrared imager, such as magnification, exposure time, etc., to obtain more detailed image information. Recording the acquisition result: and after the microscopic infrared imaging acquisition is completed, recording an acquisition result. These may include features such as shape, size, color, etc. of the abnormal region, as well as positional information in the image. Further analysis: based on the acquisition results, a deeper analysis can be performed. For example, the change of the abnormal region at different time points can be compared, or other sensor data can be combined to judge the cause of the abnormality more accurately.
Performing quality verification based on the microscopic infrared imaging acquisition result, resetting the abnormal coordinates and the abnormal region according to the quality verification result, and generating an abnormal identifier;
specifically, the quality verification: and carrying out quality verification according to microscopic infrared imaging acquisition results. This may involve evaluating indicators of sharpness, color, noise, etc. of the image to determine whether the acquired image meets quality requirements. Resetting the abnormal coordinates and the abnormal region: if the quality of the acquired image does not meet the requirements, the abnormal coordinates and the abnormal region need to be reset. This can be accomplished by comparing the previous anomaly coordinates with the anomaly area, and redetermining the anomaly coordinates and anomaly area from the currently acquired image information. Generating an anomaly identification: after resetting the abnormal coordinates and the abnormal region, an abnormal mark is generated. This identification may include information of the type of anomaly, anomaly coordinates, anomaly region, etc., to facilitate subsequent processing and tracking of the anomaly.
And recording the reset result and the identification result to generate a quality detection result.
Specifically, the reset result is recorded: recording the reset abnormal coordinates and the reset abnormal areas. This information may be stored in a database or log file for later querying and analysis. Recording an identification result: and recording the generated abnormal identification, wherein the generated abnormal identification comprises information such as abnormal type, abnormal coordinates, abnormal area and the like. Generating a quality detection result: and generating a quality detection result according to the recorded reset result and the recorded identification result. The result may be a report or warning message including descriptions of the anomaly coordinates, anomaly areas, and treatment recommendations or measures for these anomalies. It should be noted that the quality of the recorded reset results and the identification results directly affects the accuracy of the generated quality detection results. Therefore, it is necessary to ensure the authenticity and integrity of the data when recording is performed.
Further, as shown in fig. 2, the method of the present application further includes:
configuring the number of sampling points, and generating minimum sampling point distance constraint according to the field of view interval and the number of sampling points;
performing image gray processing on an image acquired by the thermal infrared imager, and performing random sampling point distribution according to the number of the sampling points and the minimum sampling point distance constraint;
sampling gray image pixel values according to the random distribution result, and recording a sampling set;
and determining a comparison benchmark by using the sampling set, executing abnormal recognition of the image acquired by the thermal infrared imager, and determining the abnormal region and the abnormal coordinates.
Specifically, the number of sampling points is configured: first, an appropriate number of sampling points needs to be selected. The number of sampling points should be determined according to factors such as application requirements, equipment performance, and size of the target object. For example, if the target object is larger, more sampling points may be required to cover the entire object. If the target object is smaller, fewer sampling points may be required to reduce the amount of computation. Generating a minimum sampling point distance constraint: next, a minimum sample point distance constraint may be generated from the field of view interval and the number of sample points. This constraint may ensure a minimum distance between each sampling point, thereby avoiding sampling points that are too dense or too sparse. The specific generation method can be determined according to actual requirements, for example, an average distance is calculated according to the size of the field of view and the number of sampling points to serve as a minimum sampling point distance constraint. Performing image gray scale processing: and then, carrying out gray processing on the image acquired by the thermal infrared imager. The gradation processing can convert a color image into a black-and-white image, thereby reducing the amount of calculation and improving the processing speed. Common gray scale processing methods include averaging, weighted averaging, thresholding, and the like. Random sample point distribution: then, random sampling point distribution is carried out according to the number of the sampling points and the minimum sampling point distance constraint. This process can randomly generate a series of pixels on the gray scale image and ensure that the distance between the pixels is not less than the minimum sampling point distance constraint. Sampling pixel values: then, the pixel value of the gray image is sampled according to the random distribution result. Specifically, a certain number of pixel points may be selected from the randomly distributed sampling points, and their pixel values may be recorded. Determining an alignment reference: and taking the sampled pixel value set as a comparison standard for abnormal recognition of the acquired image of the subsequent thermal infrared imager. Abnormality identification: and finally, based on the comparison standard, carrying out anomaly identification on the image acquired by the thermal infrared imager. This can be determined by comparing the pixel value of each point with a reference. When a pixel point having a large difference from the reference is found, it can be considered that there is an abnormality in the region corresponding to the pixel point. In this way, the abnormal region and the abnormal coordinates can be determined.
Further, the method of the present application further comprises:
calculating a sample balance value of the sampling sample by using the sampling set to obtain a balance value calculation result;
evaluating the equalization value calculation result through a standard equalization value to generate an evaluation result;
if the evaluation result is a passing result, carrying out abnormal pixel value analysis of the sampling set according to the balance value calculation result, and eliminating the abnormal pixel value based on the analysis result;
and regenerating a comparison balance value according to the eliminating result, taking the comparison balance value as a standard value, setting a tolerance window, comparing binary pixels of the gray level image through the tolerance window, and completing the abnormality identification based on the comparison result.
Specifically, sample equalization value calculation of the sampling sample is performed by the sampling set, and an equalization value calculation result is obtained: by using the sample set of the image acquired by the thermal infrared imager, a sample equalization value of the sampled sample can be calculated. Involving statistical analysis and calculation, such as calculating the average or median of each pixel value, or calculating the standard deviation of pixel values, etc. The goal of this step is to find a simple value that can represent the entire image, i.e. the equalization value. Evaluating the equalization value calculation result through a standard equalization value to generate an evaluation result: in this step, the calculated equalization value needs to be compared with a preset standard equalization value. The standard equalization value may be a known normal reference value or an average value calculated from historical data. The result of the comparison may generate an evaluation result indicating whether the sample set is within the expected range. If the evaluation result is a passing result, performing abnormal pixel value analysis of the sampling set according to the balance value calculation result, and eliminating the abnormal pixel value based on the analysis result: if the evaluation result shows that the sample set is within the normal range, the next step, i.e. analyzing the abnormal pixel values in the sample set, can be performed. This step involves an intensive study of each pixel value, for example checking whether it exceeds a preset range, or relation to other pixel values, etc. Then, based on the result of the analysis, those pixel values that are identified as abnormal may be culled. Regenerating a comparison balanced value according to a rejection result, setting a tolerant window by taking the comparison balanced value as a standard value, comparing binary pixels of a gray level image through the tolerant window, and completing anomaly identification based on the comparison result: after the outlier is removed, a new equalization value may be recalculated as the standard value. A tolerance window is then set that can receive pixel value deviations over a range. Each pixel of the greyscale image is then aligned through this forgiving window. If the value of a certain pixel is beyond the range of the tolerance window, the region corresponding to the pixel can be considered to have abnormality.
Further, as shown in fig. 3, the method of the present application further includes:
taking the number of the sampling points as uniform interval constraint, uniformly dividing the visual field interval, and determining the size of a dividing block;
acquiring a short side size and a long side size of the dividing block, and inputting the short side size and the long side size into a distance constraint network;
and executing optimizing evaluation of the constraint distance through the distance constraint network, and determining the minimum sampling point distance constraint.
Specifically, the number of sampling points is used as a uniform interval constraint, the field of view interval is uniformly divided, and the size of a division block is determined: in this step, the video field section is uniformly divided using the determined number of sampling points as a uniform section constraint. The size of each segment is determined to ensure that the pixels within each segment are sufficiently uniform. The field of view interval may be divided into rows or columns. Obtaining a short side size and a long side size of the segmented block, and inputting the short side size and the long side size into a distance constraint network: after determining the segment sizes, the short side size and the long side size of each segment are obtained and input into the distance constraint network. The distance constraint network may be a pre-trained deep learning model that is input as the short side dimension and the long side dimension of the segment and output as the minimum distance constraint between the individual sample points. By training this model, the minimum distance constraint under different sized segments can be obtained. Executing optimizing evaluation of constraint distance through the distance constraint network, and determining the minimum sampling point distance constraint: after the short side size and the long side size of the split block are input into the distance constraint network, the network outputs the corresponding minimum distance constraint. By comparing the minimum distance constraints under different size segments, the optimal distance constraint, i.e., the minimum sampling point distance constraint, can be found. This optimal distance constraint will act as a constraint on the distribution of subsequent random sample points. The whole process obtains the optimal minimum sampling point distance constraint through the joint optimization of the uniform segmentation and the distance constraint network. This constraint can ensure that the distance between sampling points is not less than this constraint in the subsequent random sampling point distribution, thereby avoiding that the sampling points are too dense or too sparse.
Further, the method of the present application further comprises:
establishing a defect characteristic set of the composite film through big data, wherein each characteristic in the defect characteristic set is provided with characteristic value mapping and range value mapping;
taking the defect feature set as a traversal convolution kernel, and executing image traversal of microscopic infrared imaging acquisition results to obtain image traversal results, wherein the image traversal results comprise adaptive compatible values and feature matching results, and the adaptive compatible values comprise positive compatible values and negative compatible values;
and calling a mapping characteristic value and a mapping range value according to the characteristic matching result, carrying out compatible adjustment on the mapping characteristic value and the mapping range value according to the adaptive compatible value, and resetting an abnormal value, an abnormal characteristic and an abnormal coordinate to finish quality verification.
Specifically, a defect feature set of the composite film is established: first, a large number of composite film samples, including defective and non-defective samples, need to be collected. These samples are then imaged using a high precision microscopic infrared imaging device and various possible defect features are extracted using image processing techniques. These features may include variations in various shapes, sizes, colors, textures, and the like. All of these features are organized into a set of features, each feature having its corresponding feature value map and range value map. Traversing the image: and taking the defect characteristic set as a traversing convolution kernel, and traversing the image result acquired by each microscopic infrared imaging. This process may detect features in the image that match the defect feature set. Acquiring an image traversing result: after traversing, an image traversal result is obtained, which includes the adaptation compatibility value and the feature matching result. The adaptive compatibility value indicates the similarity between the current feature and the standard feature, the positive compatibility value indicates the similarity between the current feature and the standard feature, and the negative compatibility value indicates the difference between the current feature and the standard feature. Adjusting the mapping characteristic value and the mapping range value: according to the feature matching result, the corresponding mapping feature value and mapping range value can be invoked. These characteristic values and range values are then compatibly adjusted according to the adaptation compatibility values. If the adaptation compatibility value of a feature is too low, it is stated that the feature may be an outlier. Resetting outliers, outlier features and outlier coordinates: after adjustment, new outliers, outlier features and outlier coordinates are obtained. And (3) finishing quality verification: through the steps, defects and anomalies in the composite film can be effectively detected, and the quality verification of the composite film is completed.
Further, the method of the present application further comprises:
performing bias evaluation of different visual field intervals in the same zero resetting interval according to the resetting abnormal coordinates;
if the deviation evaluation result meets a preset threshold, extracting the coverage commonality coordinate of the reset abnormal coordinate;
and performing speed compensation of the conveyor belt through the coverage common coordinates.
Specifically, different viewing field interval bias evaluations within the same zero point reset interval are performed according to the reset abnormal coordinates: first, the location of the reset abnormal coordinates is determined, and then the coordinates are classified, and they are classified into the same zero-point reset section. This zero-reset interval may be one or more specific field of view intervals. Next, the abnormal coordinates of the different field of view sections within these same zero-point reset sections are evaluated for bias to determine whether they have a deviation. If the deviation evaluation result meets a preset threshold value: after the abnormal coordinates of different visual field intervals in the same zero resetting interval are subjected to deviation evaluation, judging whether the deviation degree of the coordinates meets a preset threshold value or not. If satisfied, it is stated that the offset of these coordinates is large, which may affect the subsequent processing results. Extracting the coverage common coordinates of the reset abnormal coordinates: after determining the reset anomaly coordinates, their overlay commonality coordinates are extracted. These overlay commonality coordinates may refer to areas or points that are overlaid by a plurality of reset anomaly coordinates. Performing speed compensation of the conveyor belt by means of the overlay commonality coordinates: finally, the extracted covering commonality coordinates are utilized to perform speed compensation of the conveyor belt. This may be to correct the speed deviation of the conveyor belt in the zero-point reset interval to reduce the influence on the subsequent processing. The specific speed compensation strategy may need to be determined based on the actual application scenario and equipment performance, for example, it may be necessary to increase the speed of the conveyor belt at a particular location or decrease the speed of the conveyor belt. And further processing and correcting the reset abnormal coordinates to ensure the accuracy and efficiency of subsequent processing.
Further, the method of the present application further comprises:
if the deviation evaluation result cannot meet the preset threshold value, generating early warning information according to the coordinate specific value of the reset abnormal coordinate;
and carrying out early warning accumulation on the early warning information, and executing maintenance management of the conveyor belt according to the early warning accumulation result.
Specifically, if the deviation evaluation result cannot meet the preset threshold, generating early warning information according to the coordinate specific value of the reset abnormal coordinate: if the deviation evaluation result shows that the deviation degree of the abnormal coordinates does not reach the preset threshold value, the early warning information can be generated according to the coordinate specific values of the abnormal coordinates. In this process, the characteristics of the abnormal coordinates, such as position, distribution, type, etc., and their possible influence on the belt operation, need to be considered. The specific content of the early warning information can be set according to the actual situation, and for example, the information may include the type, the position, the influence degree and the like of the abnormal coordinates. And carrying out early warning accumulation on the early warning information: after the early warning information is generated, accumulated analysis is needed to be carried out on the information. This may include processing of the classification, ordering, clustering, etc. of the pre-warning information to find rules and trends therein. And performing maintenance management of the conveyor belt through early warning accumulated results: after the early warning is accumulated, the conveyor belt can be maintained and managed according to the accumulated early warning result. For example, if the amount of some type of warning information continues to increase over a period of time, then corresponding measures, such as changing components, adjusting parameters, increasing inspection, etc., may need to be taken to prevent such warning from occurring or mitigate its effects. The possible problems of the conveyor belt are found in advance through an early warning mechanism, and corresponding maintenance management measures are adopted to ensure the normal operation of the conveyor belt.
Example two
Based on the same inventive concept as the quality detection method of the composite film for medicine packaging of the foregoing embodiment, as shown in fig. 4, the present application provides a quality detection system of the composite film for medicine packaging, the system comprising:
a measurement correction length determining module 10, wherein the measurement correction length determining module 10 is used for configuring the speed of a conveyor belt, setting a length measurement sensor at a conveying node, configuring the length measurement sensor based on the stability coefficient of the conveyor belt, and determining a measurement correction length;
the visual field interval and acquisition delay determining module 20 is used for determining the visual field interval and the acquisition delay of the thermal infrared imager, setting an acquisition time node of the thermal infrared imager according to the speed of the conveyor belt, and resetting the zero point of the acquisition time node based on the metering correction length;
the wide-tolerance drift coordinate generation module 30 is used for carrying out actual stability analysis on the conveyor belt based on a zero point reset result by the wide-tolerance drift coordinate generation module 30 to generate wide-tolerance drift coordinates;
the abnormal coordinate and abnormal region identification module 40 is used for establishing a unified image coordinate system of an image acquired by the thermal infrared imager and identifying and obtaining abnormal coordinates and abnormal regions;
the micro infrared imaging acquisition result recording module 50 is used for carrying out tolerance expansion on the abnormal coordinates and the abnormal region through the tolerance drift coordinates, executing micro infrared imaging acquisition according to the tolerance expansion result and recording the micro infrared imaging acquisition result;
the abnormal identifier generation module 60, wherein the abnormal identifier generation module 60 performs quality verification based on the microscopic infrared imaging acquisition result, resets the abnormal coordinates and the abnormal region according to the quality verification result, and generates an abnormal identifier;
the quality detection result generating module 70 is configured to record the reset result and the identification result, and generate a quality detection result.
Further, the system further comprises:
the minimum sampling point distance constraint generation module is used for configuring the number of sampling points and generating minimum sampling point distance constraint according to the field of view interval and the number of sampling points;
the image gray processing module is used for carrying out image gray processing on the image acquired by the thermal infrared imager and carrying out random sampling point distribution according to the number of the sampling points and the minimum sampling point distance constraint;
the sampling set recording module is used for sampling the pixel values of the gray image according to the random distribution result and recording a sampling set;
and the abnormal region and abnormal coordinate determining module is used for determining a comparison standard by using the sampling set, executing abnormal recognition of the image acquired by the thermal infrared imager and determining the abnormal region and the abnormal coordinate.
Further, the system further comprises:
the equalization value calculation result obtaining module is used for calculating a sample equalization value of a sampling sample by using the sampling set to obtain an equalization value calculation result;
the evaluation result generation module is used for evaluating the calculation result of the equalization value through a standard equalization value to generate an evaluation result;
the abnormal pixel value eliminating module is used for executing abnormal pixel value analysis of the sampling set according to the balance value calculation result if the evaluation result is a passing result, and eliminating the abnormal pixel value based on the analysis result;
and the anomaly identification module is used for regenerating a comparison balanced value according to the eliminating result, setting a tolerant window by taking the comparison balanced value as a standard value, comparing binary pixels of the gray level image through the tolerant window, and completing anomaly identification based on the comparison result.
Further, the system further comprises:
the dividing size determining module is used for uniformly dividing the visual field interval by taking the number of the sampling points as uniform interval constraint and determining the size of a dividing block;
the distance constraint network input module is used for acquiring the short side size and the long side size of the dividing block and inputting the short side size and the long side size into a distance constraint network;
and the minimum sampling point distance constraint determining module is used for executing optimizing evaluation of constraint distance through the distance constraint network to determine minimum sampling point distance constraint.
Further, the system further comprises:
the device comprises a feature set establishing module, a feature value mapping module and a range value mapping module, wherein the feature set establishing module is used for establishing a defect feature set of the composite film through big data, and each feature in the defect feature set is provided with a feature value mapping and a range value mapping;
the image traversing result acquisition module is used for performing image traversing of microscopic infrared imaging acquisition results by taking the defect feature set as a traversing convolution kernel to obtain an image traversing result, wherein the image traversing result comprises an adaptive compatible value and a feature matching result, and the adaptive compatible value comprises a positive compatible value and a negative compatible value;
and the quality verification completion module is used for calling the mapping characteristic value and the mapping range value according to the characteristic matching result, carrying out compatible adjustment on the mapping characteristic value and the mapping range value according to the adaptive compatible value, and resetting the abnormal value, the abnormal characteristic and the abnormal coordinate so as to complete quality verification.
Further, the system further comprises:
the visual field interval bias evaluation module is used for evaluating different visual field intervals in the same zero resetting interval according to the resetting abnormal coordinates;
the coverage commonality coordinate extraction module is used for extracting the coverage commonality coordinate of the reset abnormal coordinate if the deviation evaluation result meets a preset threshold value;
and the speed compensation module is used for executing speed compensation of the conveyor belt through the coverage common coordinates.
Further, the system further comprises:
the early warning information generation module is used for generating early warning information according to the coordinate specific value of the reset abnormal coordinate if the deviation evaluation result cannot meet the preset threshold value;
and the maintenance management execution module is used for carrying out early warning accumulation on the early warning information and executing the maintenance management of the conveyor belt according to the early warning accumulation result.
The foregoing detailed description of the quality detection method of the composite film for medicine packaging will clearly be known to those skilled in the art, and the system disclosed in the examples is described more simply because it corresponds to the device disclosed in the examples, and the relevant points are described in the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. A method for quality inspection of a composite film for pharmaceutical packaging, the method comprising:
configuring the speed of a conveyor belt, setting a length measurement sensor at a conveying node, configuring the length measurement sensor based on the stability coefficient of the conveyor belt, and determining a measurement correction length;
determining a visual field interval and acquisition delay of the thermal infrared imager, setting an acquisition time node of the thermal infrared imager according to the speed of the conveyor belt, and resetting a zero point of the acquisition time node based on the metering correction length;
carrying out actual stability analysis of the conveyor belt based on the zero point resetting result to generate a tolerant drift coordinate;
establishing a unified image coordinate system of an image acquired by the thermal infrared imager, and identifying and obtaining abnormal coordinates and an abnormal region;
performing forgiving expansion on the abnormal coordinates and the abnormal region through the forgiving drift coordinates, executing microscopic infrared imaging acquisition according to the forgiving expansion result, and recording microscopic infrared imaging acquisition result;
performing quality verification based on the microscopic infrared imaging acquisition result, resetting the abnormal coordinates and the abnormal region according to the quality verification result, and generating an abnormal identifier;
and recording the reset result and the identification result to generate a quality detection result.
2. The method of claim 1, wherein the method further comprises:
configuring the number of sampling points, and generating minimum sampling point distance constraint according to the field of view interval and the number of sampling points;
performing image gray processing on an image acquired by the thermal infrared imager, and performing random sampling point distribution according to the number of the sampling points and the minimum sampling point distance constraint;
sampling gray image pixel values according to the random distribution result, and recording a sampling set;
and determining a comparison benchmark by using the sampling set, executing abnormal recognition of the image acquired by the thermal infrared imager, and determining the abnormal region and the abnormal coordinates.
3. The method of claim 2, wherein the method further comprises:
calculating a sample balance value of the sampling sample by using the sampling set to obtain a balance value calculation result;
evaluating the equalization value calculation result through a standard equalization value to generate an evaluation result;
if the evaluation result is a passing result, carrying out abnormal pixel value analysis of the sampling set according to the balance value calculation result, and eliminating the abnormal pixel value based on the analysis result;
and regenerating a comparison balance value according to the eliminating result, taking the comparison balance value as a standard value, setting a tolerance window, comparing binary pixels of the gray level image through the tolerance window, and completing the abnormality identification based on the comparison result.
4. The method of claim 2, wherein the method further comprises:
taking the number of the sampling points as uniform interval constraint, uniformly dividing the visual field interval, and determining the size of a dividing block;
acquiring a short side size and a long side size of the dividing block, and inputting the short side size and the long side size into a distance constraint network;
and executing optimizing evaluation of the constraint distance through the distance constraint network, and determining the minimum sampling point distance constraint.
5. The method of claim 1, wherein the method further comprises:
establishing a defect characteristic set of the composite film through big data, wherein each characteristic in the defect characteristic set is provided with characteristic value mapping and range value mapping;
taking the defect feature set as a traversal convolution kernel, and executing image traversal of microscopic infrared imaging acquisition results to obtain image traversal results, wherein the image traversal results comprise adaptive compatible values and feature matching results, and the adaptive compatible values comprise positive compatible values and negative compatible values;
and calling a mapping characteristic value and a mapping range value according to the characteristic matching result, carrying out compatible adjustment on the mapping characteristic value and the mapping range value according to the adaptive compatible value, and resetting an abnormal value, an abnormal characteristic and an abnormal coordinate to finish quality verification.
6. The method of claim 1, wherein the method further comprises:
performing bias evaluation of different visual field intervals in the same zero resetting interval according to the resetting abnormal coordinates;
if the deviation evaluation result meets a preset threshold, extracting the coverage commonality coordinate of the reset abnormal coordinate;
and performing speed compensation of the conveyor belt through the coverage common coordinates.
7. The method of claim 6, wherein the method further comprises:
if the deviation evaluation result cannot meet the preset threshold value, generating early warning information according to the coordinate specific value of the reset abnormal coordinate;
and carrying out early warning accumulation on the early warning information, and executing maintenance management of the conveyor belt according to the early warning accumulation result.
8. A quality inspection system for a composite film for pharmaceutical packaging, the system comprising:
the metering correction length determining module is used for configuring the speed of the conveyor belt, setting a length metering sensor at a conveying node, configuring the length metering sensor based on the stability coefficient of the conveyor belt and determining the metering correction length;
the visual field interval and acquisition delay determining module is used for determining the visual field interval and the acquisition delay of the thermal infrared imager, setting an acquisition time node of the thermal infrared imager according to the speed of the conveyor belt, and resetting the zero point of the acquisition time node based on the metering correction length;
the wide tolerance drift coordinate generation module is used for carrying out actual stability analysis on the conveyor belt based on a zero point reset result to generate wide tolerance drift coordinates;
the abnormal coordinate and abnormal region identification module is used for establishing a unified image coordinate system of an image acquired by the thermal infrared imager and identifying and obtaining abnormal coordinates and abnormal regions;
the micro infrared imaging acquisition result recording module is used for carrying out forgiving expansion on the abnormal coordinates and the abnormal region through the forgiving drift coordinates, carrying out micro infrared imaging acquisition according to the forgiving expansion results and recording micro infrared imaging acquisition results;
the abnormal identifier generation module is used for executing quality verification based on the microscopic infrared imaging acquisition result, resetting the abnormal coordinates and the abnormal region according to the quality verification result and generating an abnormal identifier;
the quality detection result generation module is used for recording the reset result and the identification result and generating a quality detection result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311420506.3A CN117169282B (en) | 2023-10-30 | 2023-10-30 | Quality detection method and system of composite film for medicine packaging |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311420506.3A CN117169282B (en) | 2023-10-30 | 2023-10-30 | Quality detection method and system of composite film for medicine packaging |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117169282A CN117169282A (en) | 2023-12-05 |
| CN117169282B true CN117169282B (en) | 2024-03-08 |
Family
ID=88939713
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311420506.3A Active CN117169282B (en) | 2023-10-30 | 2023-10-30 | Quality detection method and system of composite film for medicine packaging |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117169282B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1147889A (en) * | 1994-05-13 | 1997-04-16 | 明尼苏达州采矿制造公司 | Automated image quality control |
| CN201707309U (en) * | 2009-11-23 | 2011-01-12 | 华东交通大学 | On-line detecting device for early defects of fruit surfaces based on infrared thermal imaging technique |
| CN104597081A (en) * | 2014-12-29 | 2015-05-06 | 樊晖 | Automatic detection equipment and detection method for internal defects of plain glass |
| GB201507124D0 (en) * | 2015-04-27 | 2015-06-10 | Thermoteknix Systems Ltd | Conveyer belt monitoring system and method |
| CN109187637A (en) * | 2018-07-31 | 2019-01-11 | 中国科学院自动化研究所 | Workpiece, defect measurement method and system based on thermal infrared imager |
| CN112041662A (en) * | 2018-02-21 | 2020-12-04 | 约兰影像有限公司 | Method and system for thermal imaging of moving objects |
| CN212646548U (en) * | 2020-06-03 | 2021-03-02 | 东北林业大学 | Fruit damage nondestructive test device |
| CN113920087A (en) * | 2021-10-09 | 2022-01-11 | 东北林业大学 | Micro component defect detection system and method based on deep learning |
| CN114792323A (en) * | 2022-06-23 | 2022-07-26 | 张家港沙龙精密管业有限公司 | Steel pipe deformation detection method and system based on image processing |
-
2023
- 2023-10-30 CN CN202311420506.3A patent/CN117169282B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1147889A (en) * | 1994-05-13 | 1997-04-16 | 明尼苏达州采矿制造公司 | Automated image quality control |
| CN201707309U (en) * | 2009-11-23 | 2011-01-12 | 华东交通大学 | On-line detecting device for early defects of fruit surfaces based on infrared thermal imaging technique |
| CN104597081A (en) * | 2014-12-29 | 2015-05-06 | 樊晖 | Automatic detection equipment and detection method for internal defects of plain glass |
| GB201507124D0 (en) * | 2015-04-27 | 2015-06-10 | Thermoteknix Systems Ltd | Conveyer belt monitoring system and method |
| CN112041662A (en) * | 2018-02-21 | 2020-12-04 | 约兰影像有限公司 | Method and system for thermal imaging of moving objects |
| CN109187637A (en) * | 2018-07-31 | 2019-01-11 | 中国科学院自动化研究所 | Workpiece, defect measurement method and system based on thermal infrared imager |
| CN212646548U (en) * | 2020-06-03 | 2021-03-02 | 东北林业大学 | Fruit damage nondestructive test device |
| CN113920087A (en) * | 2021-10-09 | 2022-01-11 | 东北林业大学 | Micro component defect detection system and method based on deep learning |
| CN114792323A (en) * | 2022-06-23 | 2022-07-26 | 张家港沙龙精密管业有限公司 | Steel pipe deformation detection method and system based on image processing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117169282A (en) | 2023-12-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhou et al. | Automated visual inspection of glass bottle bottom with saliency detection and template matching | |
| CN106934800B (en) | Metal plate strip surface defect detection method and device based on YOLO9000 network | |
| US9109974B2 (en) | Tire shape inspection method and tire shape inspection apparatus | |
| Eshkevari et al. | Automatic dimensional defect detection for glass vials based on machine vision: A heuristic segmentation method | |
| US9858658B2 (en) | Defect classification using CAD-based context attributes | |
| CN113516660A (en) | Visual positioning and defect detection method and device suitable for train | |
| KR101782363B1 (en) | Vision inspection method based on learning data | |
| CN111814739B (en) | Method, device, equipment and storage medium for detecting express package volume | |
| CN103221807A (en) | Rapid processing and detection of non-uniformities in web-based materials | |
| WO2023084543A1 (en) | System and method for leveraging neural network based hybrid feature extraction model for grain quality analysis | |
| JP2015197396A (en) | Image inspection method and image inspection device | |
| CN115619708A (en) | Method for judging fault based on image recognition of oil quality change of main shaft bearing lubricating oil | |
| Felipe et al. | Vision-based liquid level detection in amber glass bottles using OpenCV | |
| CN117649404A (en) | Medicine packaging box quality detection method and system based on image data analysis | |
| CN115860510A (en) | Production efficiency analysis and evaluation method based on big data | |
| CN115439654A (en) | Method and system for finely dividing weakly supervised farmland plots under dynamic constraint | |
| CN117589109B (en) | Quality detection and control method for quartz center tray manufacturing process | |
| Börold et al. | Deep learning-based object recognition for counting car components to support handling and packing processes in automotive supply chains | |
| KR101782364B1 (en) | Vision inspection method based on learning data | |
| US9037421B2 (en) | Leak detection system for uniform vacuum packaged products | |
| CN117169282B (en) | Quality detection method and system of composite film for medicine packaging | |
| JP5804510B2 (en) | Method for inspecting the liquid content of a bottle and the bottle inspecting system | |
| CN114066810A (en) | Method and device for detecting concave-convex point defects of packaging box | |
| Maningo et al. | Crack detection with 2D wall mapping for building safety inspection | |
| CN113870328A (en) | Liquid foreign matter visual detection method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |