CN112800909A - Self-learning type tobacco shred sundry visual image detection method - Google Patents
Self-learning type tobacco shred sundry visual image detection method Download PDFInfo
- Publication number
- CN112800909A CN112800909A CN202110071378.0A CN202110071378A CN112800909A CN 112800909 A CN112800909 A CN 112800909A CN 202110071378 A CN202110071378 A CN 202110071378A CN 112800909 A CN112800909 A CN 112800909A
- Authority
- CN
- China
- Prior art keywords
- image
- tobacco shred
- sundry
- tobacco
- deep learning
- 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.)
- Pending
Links
- 241000208125 Nicotiana Species 0.000 title claims abstract description 141
- 235000002637 Nicotiana tabacum Nutrition 0.000 title claims abstract description 141
- 238000001514 detection method Methods 0.000 title claims abstract description 67
- 230000000007 visual effect Effects 0.000 title claims abstract description 22
- 238000013135 deep learning Methods 0.000 claims abstract description 54
- 238000012549 training Methods 0.000 claims abstract description 36
- 239000012535 impurity Substances 0.000 claims abstract description 15
- 238000005516 engineering process Methods 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims abstract description 8
- 238000003709 image segmentation Methods 0.000 claims abstract description 6
- 230000009466 transformation Effects 0.000 claims abstract description 5
- 238000012545 processing Methods 0.000 claims description 17
- 230000001133 acceleration Effects 0.000 claims description 11
- 230000006870 function Effects 0.000 claims description 11
- 238000004891 communication Methods 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 7
- 238000013527 convolutional neural network Methods 0.000 claims description 7
- 238000003706 image smoothing Methods 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 235000019504 cigarettes Nutrition 0.000 abstract description 11
- 238000004519 manufacturing process Methods 0.000 abstract description 10
- 238000000034 method Methods 0.000 description 15
- 238000005286 illumination Methods 0.000 description 3
- 238000013136 deep learning model Methods 0.000 description 2
- 239000000123 paper Substances 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 208000003464 asthenopia Diseases 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 239000011087 paperboard Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
- G06F18/2414—Smoothing the distance, e.g. radial basis function networks [RBFN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention provides a self-learning type tobacco shred sundry visual image detection method, which comprises the following steps: acquiring tobacco shred pictures containing different kinds of sundries, and carrying out image preprocessing on the tobacco shred pictures; extracting tobacco shred sundry images from the processed tobacco shred images by adopting an image segmentation technology, and performing transformation, rotation and shading adjustment on the tobacco shred sundry images to obtain a plurality of corresponding tobacco shred sundry images; fusing the tobacco shred standard image and the tobacco shred sundry image to form a training set image required by deep learning; establishing a deep learning algorithm model, and training the deep learning algorithm model through the training set image to predict and output a sundry target; and shooting the cut tobacco of the cut tobacco conveying belt through an industrial camera to obtain a cut tobacco image, and inputting the cut tobacco image into the trained deep learning algorithm model for impurity detection. The invention can improve the accuracy and the working efficiency of the tobacco shred sundries detection and improve the production quality of cigarettes.
Description
Technical Field
The invention relates to the technical field of cigarette production management, in particular to a self-learning type tobacco shred sundry visual image detection method.
Background
The cigarette is often inevitable to be mixed with paper or plastic packaging scraps in the transportation process in the production process of the cigarette, so that the quality of the cigarette is influenced, and therefore, the impurity detection of the cut tobacco on a production line is required. The impurity foreign matter in the pipe tobacco is look for to artifical the selecting relies on people's eye entirely, receives the subjective consciousness influence of individual, and wastes time and energy, and long-time work causes visual fatigue easily, can't guarantee work efficiency. The traditional machine vision method is greatly influenced by the shadow of the tobacco shred gaps when detecting the cigarette impurities, the colors of the paper scraps and the tobacco shreds are very close, and the detection accuracy rate needs to be improved.
The current deep learning technology is widely applied to the field of machine vision detection, but the deep learning technology has no good application case at present in the aspect of tobacco shred sundries detection application. Therefore, how to detect the sundry target in the tobacco shred image by using the deep learning technology has important significance in improving the accuracy and efficiency of target identification.
Disclosure of Invention
The invention provides a self-learning type tobacco shred sundry visual image detection method, which solves the problems of low detection efficiency and low accuracy rate of sundry detection in the conventional tobacco shreds, can improve the accuracy and the working efficiency of tobacco shred sundry detection, and improves the production quality of cigarettes.
In order to realize the following purposes, the invention provides the following technical scheme:
a self-learning type tobacco shred sundry visual image detection method comprises the following steps:
acquiring tobacco shred pictures containing different kinds of sundries, and carrying out image preprocessing on the tobacco shred pictures;
extracting tobacco shred sundry images from the processed tobacco shred images by adopting an image segmentation technology, and performing transformation, rotation and shading adjustment on the tobacco shred sundry images to obtain a plurality of corresponding tobacco shred sundry images;
fusing the tobacco shred standard image and the tobacco shred sundry image to form a training set image required by deep learning;
establishing a deep learning algorithm model, and training the deep learning algorithm model through the training set image to predict and output a sundry target;
and shooting the cut tobacco of the cut tobacco conveying belt through an industrial camera to obtain a cut tobacco image, and inputting the cut tobacco image into the trained deep learning algorithm model for impurity detection.
Preferably, the method further comprises the following steps:
and constructing the deep learning algorithm model and the convolutional neural network by adopting Python language, and calling the deep learning algorithm model through the convolutional neural network to detect sundries.
Preferably, the method further comprises the following steps:
and constructing a control interface between the industrial personal computer and the industrial camera through Python language, and controlling the operations of shooting, image stream capturing and image stream format conversion of the industrial camera through the control interface.
Preferably, the method further comprises the following steps:
and constructing a communication interface between the industrial personal computer and the PLC through a Python language, and realizing the start-stop control of the industrial personal computer on the tobacco shred conveying belt through the communication interface.
Preferably, the deep learning algorithm model adopts a Yolo V3 model and a Keras-TensorFlow deep learning GPU acceleration platform to realize GPU acceleration.
Preferably, the method further comprises the following steps:
changing the weight coefficient of the sundries-free partial image in the loss function through an experimental test on the Yolo V3 model, finding the weight coefficient most suitable for detecting sundries on the surface of the cut tobacco on the conveyor belt, and then training the training set image by using the Yolov3 model to obtain a weight file;
weight file conversion based on Tensorflow is completed through a Yolo V3 model based on Keras so as to realize impurity detection of the cut tobacco on the cut tobacco conveying belt.
Preferably, the image preprocessing of the tobacco shred picture includes:
carrying out image enhancement processing and image smoothing processing on the tobacco shred images;
and performing frame marking and frame coordinate extraction on the sundries in the tobacco shred pictures.
Preferably, the establishing a deep learning algorithm model and training the deep learning algorithm model through the training set image includes:
building a Keras-TensorFlow deep learning GPU acceleration platform;
constructing a Yolo V3 model, and defining a darknet block class, a convolution pooling block class and a convolution block class according to a set format;
and calling a Yolo V3 model and a training set image through a Keras-TensorFlow deep learning platform to perform deep learning training, and generating a weight file.
The invention provides a self-learning type tobacco shred sundry visual image detection method, which forms a training set image through a tobacco shred sundry image to train a deep learning algorithm model so as to realize the purpose of predicting and outputting sundries. The problem of the debris detects to have detection efficiency low and the rate of accuracy is not high in the current pipe tobacco is solved, can improve pipe tobacco debris detection's accuracy and work efficiency, improve the production quality of cigarette.
Drawings
In order to more clearly describe the specific embodiments of the present invention, the drawings to be used in the embodiments will be briefly described below.
FIG. 1 is a schematic diagram of a self-learning type tobacco shred sundry visual image detection method provided by the invention.
FIG. 2 is a schematic view of a tobacco shred sundry image detection system provided by the invention.
Fig. 3 is a flow chart of tobacco shred impurity detection provided by the embodiment of the invention.
Fig. 4 is a flow chart for detecting impurities in cut tobacco according to the embodiment of the present invention.
Detailed Description
In order to make the technical field of the invention better understand the scheme of the embodiment of the invention, the embodiment of the invention is further described in detail with reference to the drawings and the implementation mode.
Aiming at the problems of low efficiency and low accuracy of the conventional detection of the sundries in the tobacco shreds, the invention provides a self-learning type visual image detection method for the sundries in the tobacco shreds, which solves the problems of low detection efficiency and low accuracy of the conventional detection of the sundries in the tobacco shreds, and can improve the accuracy and the working efficiency of the detection of the sundries in the tobacco shreds and improve the production quality of cigarettes.
As shown in fig. 1, a self-learning type tobacco shred sundry visual image detection method comprises the following steps:
s1: the method comprises the steps of collecting tobacco shred pictures containing different kinds of sundries, and carrying out image preprocessing on the tobacco shred pictures.
S2: and extracting tobacco shred sundry images from the processed tobacco shred images by adopting an image segmentation technology, and performing transformation, rotation and shading adjustment on the tobacco shred sundry images to obtain a plurality of corresponding tobacco shred sundry images.
S3: and fusing the tobacco shred standard image and the tobacco shred sundry image to form a training set image required by deep learning.
S4: and establishing a deep learning algorithm model, and training the deep learning algorithm model through the training set image so as to predict and output the sundry target.
S5: and shooting the cut tobacco of the cut tobacco conveying belt through an industrial camera to obtain a cut tobacco image, and inputting the cut tobacco image into the trained deep learning algorithm model for impurity detection.
Specifically, as shown in fig. 2, the tobacco shred sundry visual image detection system comprises: the industrial personal computer is in signal connection with the industrial camera and the PLC respectively, and the PLC acquires the running state of the tobacco shred conveying belt and sends the running state to the industrial personal computer. And the industrial personal computer controls the industrial camera to photograph the cut tobacco on the cut tobacco conveyer belt when the cut tobacco conveyer belt is in the running state to obtain a cut tobacco image. And an image processing algorithm is operated on the industrial personal computer, and the tobacco shred images are subjected to image processing and sundry identification. After the system is started, the light source of the industrial camera is in a normally-on state, the state of the tobacco shred conveying belt is read by the industrial personal computer, and when the conveying belt runs, the industrial camera is controlled to photograph in real time. The system automatically captures the next image from the image stream of the industrial camera for detection after the current image is processed by the industrial personal computer. The industrial computer obtains the running state of pipe tobacco conveyer belt through the PLC controller, and after detecting debris, the PLC controller control pipe tobacco conveyer belt is shut down, realizes the auto-stop function. The industrial personal computer mainly completes the functions of reading image stream, storing sundry images, counting, converting image formats, detecting tobacco shred sundries, storing detection results, controlling communication and the like.
In an embodiment, the cut tobacco impurity detection process is as shown in fig. 3, (1) after the system is started, the industrial camera is turned on and the camera image stream capture program is called to start capturing the image stream. (2) And reading the running state of the cut tobacco conveyer belt in the PLC, transmitting the camera stream picture to a variable to be detected if the cut tobacco conveyer belt runs, and continuing to read the running state of the cut tobacco conveyer belt if the cut tobacco conveyer belt stops. (3) And after the picture is read, executing a detection program and outputting a detection result. (4) And if sundries are detected, the industrial personal computer writes a cut tobacco conveyer belt stop signal into the PLC, and the PLC drives the starting equipment to alarm and stop to wait for the inspection and processing of an operator. Checking and processing by an operator, if the sundries are really present, processing the sundries, confirming the sundries processing, and storing sundries pictures and detection results by the system; and if the operator checks that no sundries exist, carrying out false detection processing confirmation, storing a false detection picture by the system, and storing a detection result. (5) And if the sundries are not detected, continuing the next detection period and continuously executing detection. It should be noted that the image processing algorithm running on the industrial personal computer can be realized by adopting a deep learning algorithm. The system can improve the accuracy and the working efficiency of tobacco shred sundries detection and improve the production quality of cigarettes.
In practical application, an image to be detected of the cut tobacco is obtained through an industrial camera, an image segmentation algorithm is used for extracting a cut tobacco sundry image, and the image fusion algorithm is used for fusing the cut tobacco sundry image and a background image. After the tobacco shred sundry images are extracted by using an image segmentation technology, a series of operations such as transformation, rotation, shading adjustment and the like are carried out on the tobacco shred sundry images to obtain 38000 multiple tobacco shred sundry images, and normal images acquired by an industrial camera are fused with the sundry images to finally obtain training set images required by deep learning. By researching and applying the deep learning algorithm, the problem that the traditional machine vision algorithm causes system false detection due to the color change of different batches and different brands of tobacco shreds is solved. Through the application of the deep learning algorithm, the problem that the traditional visual detection technology cannot solve, such as the fact that the shape of the stem is similar to that of the cut paperboard, is solved.
The method further comprises the following steps: and constructing the deep learning algorithm model and the convolutional neural network by adopting Python language, and calling the deep learning algorithm model through the convolutional neural network to detect sundries.
The method further comprises the following steps: and constructing a control interface between the industrial personal computer and the industrial camera through Python language, and controlling the operations of shooting, image stream capturing and image stream format conversion of the industrial camera through the control interface.
The method further comprises the following steps: and constructing a communication interface between the industrial personal computer and the PLC through a Python language, and realizing the start-stop control of the industrial personal computer on the tobacco shred conveying belt through the communication interface.
In practical application, technical researches such as industrial camera control, image stream format conversion, stream capture and the like based on python language and program compiling are completed; the data acquisition and communication technology research and application based on python-snap7 are completed, the snap7 is used for realizing the communication function between the industrial personal computer and the PLC, and the development of the functions of detecting sundries shutdown, triggering detection of conveyor belt operation and the like is directly realized through the industrial personal computer.
Further, the deep learning algorithm model adopts a Yolo V3 model and a Keras-TensorFlow deep learning GPU acceleration platform to realize GPU acceleration.
In practical application, a program is realized by using a python language, the method mainly comprises the steps of constructing a core component and a deep learning model of a dark net convolutional neural network, and the GPU acceleration is realized by using a Keras-TensorFlow deep learning GPU acceleration platform.
The method further comprises the following steps: changing the weight coefficient of the sundries-free partial image in the loss function through an experimental test on the Yolo V3 model, finding the weight coefficient most suitable for detecting sundries on the surface of the cut tobacco on the conveyor belt, and then training the training set image by using the Yolov3 model to obtain a weight file; weight file conversion based on Tensorflow is completed through a Yolo V3 model based on Keras so as to realize impurity detection of the cut tobacco on the cut tobacco conveying belt.
Further, the image preprocessing is performed on the tobacco shred picture, and the image preprocessing comprises the following steps: carrying out image enhancement processing and image smoothing processing on the tobacco shred images; and performing frame marking and frame coordinate extraction on the sundries in the tobacco shred pictures.
In practical application, in a field production environment, due to the reasons that the illumination condition changes, a shot target shakes, the color of tobacco shreds is similar to that of impurities, and the like, for example, uneven illumination in the shot environment can cause the gray scale of an image to be concentrated too much, the color contrast of the detected target is too low, and the change of the color temperature of illumination can cause the color distortion of the image to further influence the detection result. For this reason, the images need to be preprocessed the first time after they are acquired. The method mainly comprises an image enhancement processing algorithm and an image smoothing processing algorithm, and the enhancement processing is carried out on the tobacco shred images on the conveying belt, so that the output images can better meet the subsequent detection requirements, and the stability and the precision of the detection are improved.
Further, the establishing a deep learning algorithm model and training the deep learning algorithm model through the training set image includes:
and a Keras-TensorFlow deep learning GPU acceleration platform is built.
And constructing a Yolo V3 model, and defining a dark net block class, a convolution pooling block class and a convolution block class according to a set format.
And calling a Yolo V3 model and a training set image through a Keras-TensorFlow deep learning platform to perform deep learning training, and generating a weight file.
In an embodiment, a self-learning type visual image detection process of tobacco shred sundries is shown in fig. 4, (1) pictures containing different types of sundries in a region to be detected on a tobacco shred conveying belt are collected and marked, and the tobacco shred sundry detection deep learning training set is manufactured. (2) And researching and analyzing a loss function of the YoloV3 deep learning model, changing the weight coefficient of an impurity-free partial image in the loss function through experimental tests, finding the weight coefficient most suitable for detecting impurities on the surface of the cut tobacco on the conveyor belt, and then training a training set by using the Yolov3 model to obtain a weight file. (3) And realizing a YooloV 3 detection model by using Keras, converting the weight file obtained in the second step to obtain a Tensorflow-based weight file, and realizing a foreign matter detection function on the cut tobacco on the tobacco-making line conveying belt by using a Keras-based YooloV 3 detection model.
Therefore, the invention provides a self-learning tobacco shred sundry visual image detection method, which forms a training set image through a tobacco shred sundry image to train a deep learning algorithm model and realize the prediction of an output sundry target. The problem of the debris detects to have detection efficiency low and the rate of accuracy is not high in the current pipe tobacco is solved, can improve pipe tobacco debris detection's accuracy and work efficiency, improve the production quality of cigarette.
The construction, features and functions of the present invention have been described in detail with reference to the embodiments shown in the drawings, but the present invention is not limited to the embodiments shown in the drawings, and all equivalent embodiments modified or modified by the spirit and scope of the present invention should be protected without departing from the spirit of the present invention.
Claims (8)
1. A self-learning type tobacco shred sundry visual image detection method is characterized by comprising the following steps:
acquiring tobacco shred pictures containing different kinds of sundries, and carrying out image preprocessing on the tobacco shred pictures;
extracting tobacco shred sundry images from the processed tobacco shred images by adopting an image segmentation technology, and performing transformation, rotation and shading adjustment on the tobacco shred sundry images to obtain a plurality of corresponding tobacco shred sundry images;
fusing the tobacco shred standard image and the tobacco shred sundry image to form a training set image required by deep learning;
establishing a deep learning algorithm model, and training the deep learning algorithm model through the training set image to predict and output a sundry target;
and shooting the cut tobacco of the cut tobacco conveying belt through an industrial camera to obtain a cut tobacco image, and inputting the cut tobacco image into the trained deep learning algorithm model for impurity detection.
2. The self-learning type tobacco shred sundry visual image detection method according to claim 1, further comprising the following steps:
and constructing the deep learning algorithm model and the convolutional neural network by adopting Python language, and calling the deep learning algorithm model through the convolutional neural network to detect sundries.
3. The self-learning type tobacco shred sundry visual image detection method according to claim 2, further comprising the following steps:
and constructing a control interface between the industrial personal computer and the industrial camera through Python language, and controlling the operations of shooting, image stream capturing and image stream format conversion of the industrial camera through the control interface.
4. The self-learning type tobacco shred sundry visual image detection method according to claim 3, further comprising the following steps:
and constructing a communication interface between the industrial personal computer and the PLC through a Python language, and realizing the start-stop control of the industrial personal computer on the tobacco shred conveying belt through the communication interface.
5. The self-learning type visual image detection method for the sundries in the cut tobaccos according to claim 4, wherein a Yolo V3 model is adopted in the deep learning algorithm model, and a Keras-TensorFlow deep learning GPU acceleration platform is adopted to achieve GPU acceleration.
6. The self-learning type tobacco shred sundry visual image detection method according to claim 5, further comprising the following steps:
changing the weight coefficient of the sundries-free partial image in the loss function through an experimental test on the Yolo V3 model, finding the weight coefficient most suitable for detecting sundries on the surface of the cut tobacco on the conveyor belt, and then training the training set image by using the Yolov3 model to obtain a weight file;
weight file conversion based on Tensorflow is completed through a Yolo V3 model based on Keras so as to realize impurity detection of the cut tobacco on the cut tobacco conveying belt.
7. The self-learning type tobacco shred sundry visual image detection method according to claim 6, wherein the image preprocessing of the tobacco shred picture comprises the following steps:
carrying out image enhancement processing and image smoothing processing on the tobacco shred images;
and performing frame marking and frame coordinate extraction on the sundries in the tobacco shred pictures.
8. The self-learning type tobacco shred sundry visual image detection method according to claim 7, wherein the establishing of the deep learning algorithm model and the training of the deep learning algorithm model through the training set image comprise:
building a Keras-TensorFlow deep learning GPU acceleration platform;
constructing a Yolo V3 model, and defining a darknet block class, a convolution pooling block class and a convolution block class according to a set format;
and calling a Yolo V3 model and a training set image through a Keras-TensorFlow deep learning platform to perform deep learning training, and generating a weight file.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110071378.0A CN112800909A (en) | 2021-01-19 | 2021-01-19 | Self-learning type tobacco shred sundry visual image detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110071378.0A CN112800909A (en) | 2021-01-19 | 2021-01-19 | Self-learning type tobacco shred sundry visual image detection method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112800909A true CN112800909A (en) | 2021-05-14 |
Family
ID=75810578
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110071378.0A Pending CN112800909A (en) | 2021-01-19 | 2021-01-19 | Self-learning type tobacco shred sundry visual image detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112800909A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113421235A (en) * | 2021-06-17 | 2021-09-21 | 中国电子科技集团公司第四十一研究所 | Cigarette positioning device and method based on deep learning |
| WO2023047157A1 (en) * | 2021-09-23 | 2023-03-30 | Sensetime International Pte. Ltd. | Image generating methods and apparatuses, and detecting methods and apparatuses |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160071264A1 (en) * | 2014-09-06 | 2016-03-10 | RaPID Medical Technologies, LLC | Medical image dectection system and method |
| CN109297969A (en) * | 2018-12-05 | 2019-02-01 | 上海庚葵贸易有限公司 | A kind of tobacco leaf assembly line method for detecting impurities and system |
| CN111815605A (en) * | 2020-07-09 | 2020-10-23 | 成都协创信和科技有限公司 | Sleeper defect detection method based on step-by-step deep learning and storage medium |
| CN112189877A (en) * | 2020-10-13 | 2021-01-08 | 燕山大学 | On-line detection method for tobacco shred impurities in tobacco production line |
-
2021
- 2021-01-19 CN CN202110071378.0A patent/CN112800909A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160071264A1 (en) * | 2014-09-06 | 2016-03-10 | RaPID Medical Technologies, LLC | Medical image dectection system and method |
| CN109297969A (en) * | 2018-12-05 | 2019-02-01 | 上海庚葵贸易有限公司 | A kind of tobacco leaf assembly line method for detecting impurities and system |
| CN111815605A (en) * | 2020-07-09 | 2020-10-23 | 成都协创信和科技有限公司 | Sleeper defect detection method based on step-by-step deep learning and storage medium |
| CN112189877A (en) * | 2020-10-13 | 2021-01-08 | 燕山大学 | On-line detection method for tobacco shred impurities in tobacco production line |
Non-Patent Citations (2)
| Title |
|---|
| 刘迪清: "基于机器视觉的杂物识别的关键技术研究与应用", 南京航空航天大学, no. 2013, pages 1 - 46 * |
| 洪金华;忻惠琴;陆海华;刘瑞东;舒梦;陈良;陈赞赞;: "基于YOLOV3模型的卷烟厂烟虫识别方法", 烟草科技, no. 09, pages 77 - 83 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113421235A (en) * | 2021-06-17 | 2021-09-21 | 中国电子科技集团公司第四十一研究所 | Cigarette positioning device and method based on deep learning |
| WO2023047157A1 (en) * | 2021-09-23 | 2023-03-30 | Sensetime International Pte. Ltd. | Image generating methods and apparatuses, and detecting methods and apparatuses |
| AU2021240272A1 (en) * | 2021-09-23 | 2023-04-06 | Sensetime International Pte. Ltd. | Image generating methods and apparatuses, and detecting methods and apparatuses |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107392896B (en) | A kind of Wood Defects Testing method and system based on deep learning | |
| CN109389161B (en) | Garbage identification evolutionary learning method, device, system and medium based on deep learning | |
| CN112800909A (en) | Self-learning type tobacco shred sundry visual image detection method | |
| CN107402220A (en) | A kind of selective laser fusing shaping powdering quality vision online test method and system | |
| WO2020232816A1 (en) | Quality inspection method for target object, and edge computing device | |
| CN110646425B (en) | Tobacco leaf online auxiliary grading method and system | |
| CN112730445A (en) | Tobacco shred sundry visual image detection system | |
| CN115065708B (en) | Industrial Internet of things system based on machine vision detection and control method thereof | |
| CN104048966B (en) | The detection of a kind of fabric defect based on big law and sorting technique | |
| CN111127448A (en) | Model and method for detecting air spring fault based on isolated forest | |
| CN111507325B (en) | Industrial visual OCR recognition system and method based on deep learning | |
| CN115661057A (en) | Industrial nondestructive testing system and method based on cloud edge cooperation and deep learning | |
| CN108664886A (en) | A kind of fast face recognition method adapting to substation's disengaging monitoring demand | |
| CN103020635B (en) | Sintering belt temperature checking method and system in rotary kiln | |
| CN113469938B (en) | Pipe gallery video analysis method and system based on embedded front-end processing server | |
| CN102680488B (en) | Device and method for identifying massive agricultural product on line on basis of PCA (Principal Component Analysis) | |
| Kähler et al. | Anomaly detection for industrial surface inspection: Application in maintenance of aircraft components | |
| CN110618129A (en) | Automatic power grid wire clamp detection and defect identification method and device | |
| CN103984967B (en) | A kind of automatic checkout system being applied to Commercial goods labels detection and automatic testing method | |
| CN214844880U (en) | Tobacco shred sundry visual image detection system | |
| CN115564031A (en) | Detection network for glass defect detection | |
| CN104964989B (en) | A kind of method and system for photoelectricity module vision-based detection | |
| CN115379197A (en) | Method and device for detecting lens surface of camera product on production line | |
| CN104897688B (en) | The accurate processing method of image and system in a kind of chemical fiber paper tube outward appearance detection process | |
| CN112733676A (en) | Method for detecting and identifying garbage in elevator based on deep learning |
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 |