CN115240122A - Air preheater area identification method based on deep reinforcement learning - Google Patents
Air preheater area identification method based on deep reinforcement learning Download PDFInfo
- Publication number
- CN115240122A CN115240122A CN202211158753.6A CN202211158753A CN115240122A CN 115240122 A CN115240122 A CN 115240122A CN 202211158753 A CN202211158753 A CN 202211158753A CN 115240122 A CN115240122 A CN 115240122A
- Authority
- CN
- China
- Prior art keywords
- area
- air preheater
- identification
- model
- running state
- 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
- 238000000034 method Methods 0.000 title claims abstract description 23
- 230000002787 reinforcement Effects 0.000 title claims abstract description 14
- 238000001514 detection method Methods 0.000 claims abstract description 26
- 238000012706 support-vector machine Methods 0.000 claims abstract description 21
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims abstract description 14
- 230000009466 transformation Effects 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 28
- 238000012360 testing method Methods 0.000 claims description 11
- 238000012544 monitoring process Methods 0.000 claims description 10
- 238000007781 pre-processing Methods 0.000 claims description 7
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000009432 framing Methods 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 4
- 230000008901 benefit Effects 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims description 3
- 238000001931 thermography Methods 0.000 claims description 3
- 230000000007 visual effect Effects 0.000 claims description 3
- 230000008569 process Effects 0.000 claims description 2
- 238000002485 combustion reaction Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 2
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 239000003546 flue gas Substances 0.000 description 2
- 239000002699 waste material Substances 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008021 deposition Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003647 oxidation Effects 0.000 description 1
- 238000007254 oxidation reaction Methods 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000002407 reforming Methods 0.000 description 1
- 239000007921 spray Substances 0.000 description 1
- 238000005507 spraying Methods 0.000 description 1
- 239000010959 steel 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
- G06V20/40—Scenes; Scene-specific elements in video content
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- 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
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E20/00—Combustion technologies with mitigation potential
- Y02E20/34—Indirect CO2mitigation, i.e. by acting on non CO2directly related matters of the process, e.g. pre-heating or heat recovery
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to an air preheater area identification method based on deep reinforcement learning, which comprises the steps of selecting an identification area of an operation state image in a video stream, selecting a template frame, sending the operation state image in the video stream into a trained optimal support vector machine model for classification, detecting the identification area of the operation state image obtained by classification by using a NanoDet model to obtain required feature points, carrying out feature point affine matching by using the template frame as the output result of the NanoDet model, and completing the identification of a grid area of a detection frame, thereby identifying the grid area of an air preheater rotor. The method uses the support vector machine model to classify the video stream to obtain the optimal detection frame, uses the NanoDet model to detect the optimal detection frame, finally uses affine transformation to obtain the specific positions of other areas in the running state image, and can well complete the task of positioning the area of the air preheater.
Description
Technical Field
The invention relates to the field of image recognition of air preheaters, in particular to a region recognition method of an air preheater based on deep reinforcement learning.
Background
The air heater is one of the main parts of power plant boiler, and its main function is for utilizing the waste flue gas that boiler combustion discharged to preheat the air that is about to get into the boiler, and the waste flue gas that discharges from furnace passes through the economizer and gets into air heater, carries out the heat exchange with the air that is about to get into boiler combustion in air heater to heat the air, improved the combustion efficiency of boiler. Because there are a lot of gaps in the air heater heat accumulation component, can become the ferrometal iron oxidation under the condition of high temperature oxygen boosting when the heat accumulation component steel, will take place "afterburning" this moment, air heater will take place extensive conflagration, at this moment if do not in time spray water and put out a fire and handle conflagration hot spot, can cause huge loss. However, if the deposited combustible materials do not form fire hot spots, water spraying fire extinguishing measures are taken, boiler shutdown is caused, and loss is serious. Therefore, measures are taken in time at the initial stage of formation of the fire hot spot to prevent the occurrence of such an event, and therefore a detection device for the hot spot needs to be installed, so that the hot spot is detected at the initial stage of fire detection, and the detection device is effective on the premise that the area of the air preheater can be effectively distinguished. The prior art only has a method for detecting hot spots of an air preheater, such as CN201510651471, but no document discloses an air preheater area identification and location technology.
With the continuous breakthrough of the related technologies in the field of computer vision, the fault diagnosis technology based on computer vision is developed rapidly, which provides new possibility for hot spot detection of the air preheater. The background noise of the hot spot detection environment of the air preheater is high, so that the identification of the area of the air preheater by using computer vision is difficult.
Disclosure of Invention
The invention mainly solves the technical problem that an area identification method of an air preheater based on deep reinforcement learning is used, under the premise of not reforming original facilities in the air preheater, thermal infrared imagers are installed at the cold end and the hot end of the air preheater to monitor the operation of an air preheater rotor in real time, and the situation that model redundancy is increased due to repeated detection is avoided according to the operation characteristics of the air preheater rotor: intercepting a video stream sample according to the rotation time of the air preheater rotor for one circle, obtaining the running state image of the air preheater rotor in the time period through video framing, and then processing the running state image to realize the grid area identification of the air preheater rotor.
In order to achieve the purpose, the invention adopts the following technical scheme: an air preheater area identification method based on deep reinforcement learning comprises the following steps:
s1, video stream acquisition and running state image preprocessing;
s2, intercepting the identification area of the preprocessed running state image by using a cross feature frame to obtain an interception image of the identification area, and performing affine transformation;
s3, dividing the intercepted image of the identification region into a positive sample and a negative sample according to the completeness of the intercepted characteristics of the identification region for training a support vector machine model;
s4, classifying all running state images and labeling positive samples: traversing all the running state images of the air preheater rotor, putting the recognition region intercepting graphs of the running state images into an optimal support vector machine model for classification, and judging the recognition region intercepting graphs of the positive samples as a training data set of a NanoDet model;
s5, using the marked training data set for training a NanoDet model;
and S6, inputting the video stream of the air preheater rotor monitored by the thermal infrared imager at the rest time into the trained optimal support vector machine model and the NanoDet model for testing, and identifying the grid area of the air preheater rotor.
Further preferably, the step S6 specifically includes the following steps:
s6.1, intercepting an identification area of the running state image in the video stream to obtain an identification area interception image, and performing affine transformation;
s6.2, inputting the captured image of the identification area into an optimal support vector machine model for classification, executing the step S6.3 if the captured image is a positive sample, and turning to the step S6.1 to continue detection until the video stream is finished if the captured image is a negative sample;
s6.3, inputting the captured image of the identification area of the positive sample into a NanoDet model, acquiring specific coordinates of the cross area, and obtaining top coordinates of the upper left part, the lower left part and the upper right part of the cross area as feature points;
and S6.4, selecting an identification area interception image with optimal feature integrity from the positive sample in the step S3, setting an original frame of the interception image as a template frame, drawing a grid template of the rotor according to the template frame, and performing feature point affine matching on the output result of the NanoDet model, so that the grid template of the template frame is matched with the detection frame, an identification area of the detection frame corresponding to the identification area of the template frame is obtained, the grid area identification of the detection frame is completed, and the grid area identification of the air preheater rotor is realized.
Preferably, in the step S1, thermal infrared imagers are respectively installed at the cold and hot ends of the air preheater and are used for monitoring the operation condition of the rotor of the air preheater in real time and collecting video streams; the monitoring of the working condition of the air preheater becomes more visual due to the advantages of infrared thermal imaging, the monitoring video stream of the air preheater rotor running for a circle is transmitted to the local end through data transmission, the running state image of the air preheater rotor is obtained through video framing, and the running state image is subjected to Gaussian filtering and denoising through the image preprocessing module.
More preferably, in step S2, the extracted identification area is a cross area in the middle of the operation state image.
Further preferably, in step S3, the RBF radial basis function is selected for the support vector machine kernel, and the regularization parameter C and the kernel coefficient gamma are determined by a grid search method.
Further preferably, in step S4, frame skipping detection is performed for 10 frames each time a positive sample identification area is determined, if the positive sample identification area is a negative sample, the next frame is determined, the training data set is labeled, a cross area of the identification area is labeled, and a vertex of the cross area is a feature point.
The invention has the beneficial effects that: the method comprises the steps of monitoring the running state of an air preheater rotor in real time based on the actual running environment of the air preheater, effectively obtaining state information of the air preheater on the premise of not changing original equipment of the air preheater, classifying running state images through a support vector machine, detecting the running state images through a deep neural network NanoDet model, determining other regions by affine transformation, finely recognizing grid regions of the air preheater rotor, and facilitating quick positioning of subsequent rotor dust deposition faults. The invention only needs 0.2s for detecting each image in the CPU environment, and meets the requirement of hot spot detection of the air preheater of the power plant. The invention has high recognition rate of each area under the running state of the air preheater, and the precision reaches 90 percent.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
The present invention will now be described more fully hereinafter with reference to the accompanying drawings.
As shown in fig. 1, an air preheater area identification method based on deep reinforcement learning includes the following steps:
s1, video stream acquisition and running state image preprocessing: the method comprises the following steps that thermal infrared imagers are respectively installed at the cold and hot ends of an air preheater and used for monitoring the operation condition of a rotor of the air preheater in real time and collecting video streams; the advantage of infrared thermal imaging makes the monitoring of air heater operating mode become more visual, through data transmission, with the monitoring video stream transmission of air heater rotor operation a week to local end, obtains the running state image of this period air heater rotor through the video framing, rethread image preprocessing module carries out gaussian filtering and removes noise to the running state image.
And S2, intercepting the identification area of the preprocessed running state image by using the cross feature frame to obtain an identification area interception image, and performing affine transformation. The intercepted identification area is a cross area in the middle of the running state image, the principle of the intercepted identification area is that the number of features in a cross feature frame is judged manually, and the features are more and the noise is less. The affine transformation rotates 45 degrees clockwise, and the affine transformation is used for standing the identification region in the vertical direction and is beneficial to extracting the feature point position information of the identification region by a subsequent NanoDet model.
S3, dividing the recognition area intercepting graph into a positive sample and a negative sample for training a support vector machine model according to the completeness of the intercepted recognition area features;
and selecting 41 recognition area capture images as positive samples according to the integrity of the features of the recognition areas, and selecting 41 recognition area capture images as negative samples. And taking 2/3 of the positive and negative samples as a training set, and taking the rest positive and negative samples as a testing set. The kernel of the support vector machine selects RBF radial basis function, and the regularization parameter C and the kernel coefficient gamma are determined by a grid search method; and evaluating the parameter selection of the support vector machine by using the test effect of the test set so as to obtain the optimal support vector machine model. The optimal support vector machine model can remove a large number of irrelevant negative samples when traversing the image frame, the obtained positive sample data set is a simplified small sample, and irrelevant training time is greatly shortened for subsequent deep reinforcement learning of the NanoDet network model.
S4, classifying all running state images and labeling positive samples: traversing all the running state images of the air preheater rotor, putting the identification region captured image of the running state image into an optimal support vector machine model for classification, judging that the identification region captured image of a positive sample is used as a training data set of a NanoDet model, and performing frame skipping detection of 10 frames when judging a positive sample identification region, wherein the purpose of frame skipping is to avoid repeated detection of the identification region of a cross region, if the positive sample is a negative sample, performing next frame judgment, marking the training data set, marking the cross region of the identification region, and the vertex of the cross region is a feature point; the coordinate information of the positions of the characteristic points in the cross area is selected according to the principle that three points which are not on a straight line determine a surface.
S5, using the marked training data set for training a NanoDet model;
the training data set is a small sample selected from 1981 original data sets framed by a video stream of one circle of the whole rotor, and the training set and the test set are divided. And preprocessing the training data set by using Labelimg to prepare the training data set in the VOC2007 format. Configuring the running environment of the Nanodet, modifying the folders stored by training, the category number of the training, the category names of the training, the sizes and the paths of the training set and the test set, and the size of the training batch. The NanoDet model was trained with the set of well-done training sets with a learning rate of 0.001 and a number of iterations set to 300. And testing by using the test set after the training is finished, and obtaining the NanoDet model after the test is qualified.
And S6, inputting the video stream of the air preheater rotor monitored by the thermal infrared imager at the rest time into the trained optimal support vector machine model and the NanoDet model for testing, and identifying the grid area of the air preheater rotor.
S6.1, intercepting an identification area of the running state image in the video stream to obtain an identification area intercepting image, and performing affine transformation;
s6.2, inputting the recognition area captured image into an optimal support vector machine model for classification, if the recognition area captured image is a positive sample, executing the step S6.3, and if the recognition area captured image is a negative sample, turning to the step S6.1 to continue detection until the video stream is finished;
s6.3, inputting the captured image of the identification area of the positive sample into a NanoDet model, acquiring specific coordinates of the cross area, and obtaining vertex coordinates of the upper left, the lower left and the upper right of the cross area as feature points;
and S6.4, selecting an identification region interception image with the optimal feature integrity degree from the positive sample in the step S3, setting an original frame of the interception image as a template frame, drawing a grid template of the rotor according to the template frame, and performing feature point affine matching on the output result of the NanoDet model, so that the grid template of the template frame is matched with the detection frame, the identification region of the detection frame corresponding to the identification region of the template frame is obtained, the grid region identification of the detection frame is completed, and the grid region identification of the air preheater rotor is realized.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes performed by the present specification and drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (6)
1. An air preheater area identification method based on deep reinforcement learning is characterized by comprising the following steps:
s1, video stream acquisition and running state image preprocessing;
s2, intercepting the identification area of the preprocessed running state image by using a cross feature frame to obtain an interception image of the identification area, and performing affine transformation;
s3, dividing the intercepted image of the identification region into a positive sample and a negative sample according to the completeness of the intercepted characteristics of the identification region for training a support vector machine model;
s4, classifying all running state images and labeling positive samples: traversing the running state images of all air preheater rotors, putting the recognition region intercepting graphs of the running state images into an optimal support vector machine model for classification, and judging that the recognition region intercepting graphs of the positive samples are used as a training data set of a NanoDet model;
s5, using the marked training data set for training a NanoDet model;
and S6, inputting the video stream of the air preheater rotor monitored by the thermal infrared imager at the rest time into the trained optimal support vector machine model and the NanoDet model for testing, and identifying the grid area of the air preheater rotor.
2. The deep reinforcement learning-based air preheater area identification method according to claim 1, wherein the specific process of step S6 is as follows:
s6.1, intercepting an identification area of the running state image in the video stream to obtain an identification area intercepting image, and performing affine transformation;
s6.2, inputting the recognition area captured image into an optimal support vector machine model for classification, if the recognition area captured image is a positive sample, executing the step S6.3, and if the recognition area captured image is a negative sample, turning to the step S6.1 to continue detection until the video stream is finished;
s6.3, inputting the captured image of the identification area of the positive sample into a NanoDet model, acquiring specific coordinates of the cross area, and obtaining top coordinates of the upper left part, the lower left part and the upper right part of the cross area as feature points;
and S6.4, selecting an identification region interception image with the optimal feature integrity degree from the positive sample in the step S3, setting an original frame of the interception image as a template frame, drawing a grid template of the rotor according to the template frame, and performing feature point affine matching on the output result of the NanoDet model, so that the grid template of the template frame is matched with the detection frame, the identification region of the detection frame corresponding to the identification region of the template frame is obtained, the grid region identification of the detection frame is completed, and the grid region identification of the air preheater rotor is realized.
3. The deep reinforcement learning-based air preheater area identification method according to claim 1, wherein in the step S1, thermal infrared imagers are respectively mounted at the cold and hot ends of the air preheater for monitoring the operation condition of an air preheater rotor in real time and acquiring video streams; the monitoring of the working condition of the air preheater becomes more visual due to the advantages of infrared thermal imaging, the monitoring video stream of the air preheater rotor running for a circle is transmitted to the local end through data transmission, the running state image of the air preheater rotor is obtained through video framing, and the running state image is subjected to Gaussian filtering and denoising through the image preprocessing module.
4. The method for identifying the area of the air preheater based on the deep reinforcement learning as claimed in claim 1, wherein the identified area intercepted in the step S2 is a cross area in the middle of the running state image.
5. The method for identifying the area of the air preheater based on the deep reinforcement learning as claimed in claim 1, wherein in the step S3, the RBF radial basis function is selected as the kernel of the support vector machine, and the regularization parameter C and the kernel coefficient gamma are determined by a grid search method.
6. The method for identifying the area of the air preheater based on the deep reinforcement learning as claimed in claim 1, wherein in the step S4, frame skipping detection of 10 frames is performed every time a positive sample identification area is determined, if the positive sample identification area is a negative sample, the next frame determination is performed, the training data set is labeled, a cross area of the identification area is labeled, and the vertex of the cross area is a feature point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211158753.6A CN115240122A (en) | 2022-09-22 | 2022-09-22 | Air preheater area identification method based on deep reinforcement learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202211158753.6A CN115240122A (en) | 2022-09-22 | 2022-09-22 | Air preheater area identification method based on deep reinforcement learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN115240122A true CN115240122A (en) | 2022-10-25 |
Family
ID=83667191
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202211158753.6A Pending CN115240122A (en) | 2022-09-22 | 2022-09-22 | Air preheater area identification method based on deep reinforcement learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115240122A (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104881662A (en) * | 2015-06-26 | 2015-09-02 | 北京畅景立达软件技术有限公司 | Single-image pedestrian detection method |
CN105424330A (en) * | 2014-08-29 | 2016-03-23 | 北京航空航天大学 | Lorry shaft end bolt fault detecting method and lorry shaft end bolt fault detecting device |
CN108171244A (en) * | 2016-12-07 | 2018-06-15 | 北京深鉴科技有限公司 | Object identifying method and system |
CN113362302A (en) * | 2021-06-03 | 2021-09-07 | 西南交通大学 | Fault detection method of subway train electric box cover based on image recognition |
CN114120300A (en) * | 2021-10-20 | 2022-03-01 | 前锦网络信息技术(上海)有限公司 | Picture correction method and device |
CN114511527A (en) * | 2022-01-24 | 2022-05-17 | 安徽大学 | Textile spinning cake forming defect detection method based on expanded local binary pattern |
-
2022
- 2022-09-22 CN CN202211158753.6A patent/CN115240122A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105424330A (en) * | 2014-08-29 | 2016-03-23 | 北京航空航天大学 | Lorry shaft end bolt fault detecting method and lorry shaft end bolt fault detecting device |
CN104881662A (en) * | 2015-06-26 | 2015-09-02 | 北京畅景立达软件技术有限公司 | Single-image pedestrian detection method |
CN108171244A (en) * | 2016-12-07 | 2018-06-15 | 北京深鉴科技有限公司 | Object identifying method and system |
CN113362302A (en) * | 2021-06-03 | 2021-09-07 | 西南交通大学 | Fault detection method of subway train electric box cover based on image recognition |
CN114120300A (en) * | 2021-10-20 | 2022-03-01 | 前锦网络信息技术(上海)有限公司 | Picture correction method and device |
CN114511527A (en) * | 2022-01-24 | 2022-05-17 | 安徽大学 | Textile spinning cake forming defect detection method based on expanded local binary pattern |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109886298B (en) | Weld quality detection method based on convolutional neural network | |
EP2776815B1 (en) | System and method for improving automated defect detection user interface | |
CN113283344B (en) | Mining conveyor belt deviation detection method based on semantic segmentation network | |
CN109658387A (en) | The detection method of the pantograph carbon slide defect of power train | |
CN109978835B (en) | Online assembly defect identification system and method thereof | |
CN112697798B (en) | Infrared image-oriented diagnosis method and device for current-induced thermal defects of power transformation equipment | |
CN107749058A (en) | A kind of machine vision detection method and system of boiler tubing surface defect | |
CN110473806A (en) | The deep learning identification of photovoltaic cell sorting and control method and device | |
CN117250208B (en) | Machine vision-based nano-imprint wafer defect accurate detection system and method | |
CN114677362B (en) | Surface defect detection method based on improved YOLOv5 | |
CN106599890A (en) | Transformer substation patrol robot digital type instrument identification algorithm | |
CN104536396A (en) | Soft measurement modeling method used in cement raw material decomposing process in decomposing furnace | |
CN115240122A (en) | Air preheater area identification method based on deep reinforcement learning | |
CN113487570A (en) | High-temperature continuous casting billet surface defect detection method based on improved yolov5x network model | |
CN112927222B (en) | Method for realizing multi-type photovoltaic array hot spot detection based on hybrid improved Faster R-CNN | |
CN111028250A (en) | Real-time intelligent cloth inspecting method and system | |
CN117252840B (en) | Photovoltaic array defect elimination evaluation method and device and computer equipment | |
Kulkarni et al. | Machine learning based classification of welded components | |
CN111708830A (en) | Photovoltaic power station photovoltaic panel pollutant intelligent detection auxiliary system and method based on image processing | |
CN112686886A (en) | Power inspection system and equipment fault diagnosis method thereof | |
CN111798441A (en) | Photovoltaic power station photovoltaic panel pollutant detection method based on image processing | |
Kim et al. | CECvT: Initial Diagnosis of Anomalies in Thermal Images | |
TWI765364B (en) | Artificial intelligence defect image classification method and system thereof | |
KR102525249B1 (en) | Method, device and system for monitoring and analyzing anomalies in photovoltaic power plants through artificial intelligence-based image processing | |
CN112465784B (en) | Metro clamp appearance abnormality detection method |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20221025 |