CN107679581B - Method for processing gas flow distribution based on characteristic values of infrared image pixel matrix - Google Patents

Method for processing gas flow distribution based on characteristic values of infrared image pixel matrix Download PDF

Info

Publication number
CN107679581B
CN107679581B CN201710991544.2A CN201710991544A CN107679581B CN 107679581 B CN107679581 B CN 107679581B CN 201710991544 A CN201710991544 A CN 201710991544A CN 107679581 B CN107679581 B CN 107679581B
Authority
CN
China
Prior art keywords
gas flow
infrared
image
processing
infrared image
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
Application number
CN201710991544.2A
Other languages
Chinese (zh)
Other versions
CN107679581A (en
Inventor
石琳
韩博
曹富军
张景
丁根远
温有斌
马祥
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Inner Mongolia University of Science and Technology
Original Assignee
Inner Mongolia University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Inner Mongolia University of Science and Technology filed Critical Inner Mongolia University of Science and Technology
Priority to CN201710991544.2A priority Critical patent/CN107679581B/en
Publication of CN107679581A publication Critical patent/CN107679581A/en
Application granted granted Critical
Publication of CN107679581B publication Critical patent/CN107679581B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10048—Infrared image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G06T2207/20032—Median filtering

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Quality & Reliability (AREA)
  • Radiation Pyrometers (AREA)
  • Investigating Or Analysing Materials By Optical Means (AREA)

Abstract

The invention discloses a method for processing gas flow distribution based on characteristic values of an infrared image pixel matrix, and belongs to the technical field of blast furnace detection. The method comprises the following steps: data acquisition and processing; processing an infrared image; solving the eigenvalue and the eigenvector of the pixel matrix of each frame of image; and classifying the infrared images by using the size of the characteristic value. The method for processing the coal gas flow distribution based on the characteristic values of the infrared image pixel matrix classifies the large-sample blast furnace coal gas flow infrared images, finds out the dynamic change characteristic of the coal gas flow distribution in the production process of the blast furnace, provides help for further researching the relation between the coal gas flow change and the coal gas utilization rate, and further realizes the real-time monitoring of the coal gas utilization rate.

Description

Method for processing gas flow distribution based on characteristic values of infrared image pixel matrix
Technical Field
The invention particularly relates to a method for processing gas flow distribution based on characteristic values of an infrared image pixel matrix, and belongs to the technical field of blast furnace detection.
Background
At present, scholars mainly detect and analyze gas flow distribution through a blast furnace detection device, and scholars also establish a mathematical model by using infrared image characteristics of gas flow to analyze the distribution rule of the gas flow of the blast furnace, but the data volume of the infrared image of the gas flow used by the method is very small, has no statistical regularity and cannot comprehensively represent complex blast furnace conditions. For example, what is the distribution characteristics of the gas flow when the blast furnace is running forward, distributing and blowing; what is the coal gas flow distribution characteristics in case of abnormal furnace conditions. Therefore, a large amount of gas flow infrared image data (production data all year round or a period of time) in the production process of the blast furnace are classified and identified, and the search for the statistical rule of the production of the blast furnace is crucial to the production regulation and control of the blast furnace.
The prior art for studying gas flow distribution is as follows:
through the operation experience of the blast furnace for many years and the development characteristics of the gas flow, students divide the distribution state of the gas flow into four types, namely flat distribution, edge development, center development, common development of center and edge gas flows and the like. The technology only analyzes the distribution characteristics of the gas flow of the blast furnace on the theoretical basis, and does not analyze the distribution of the gas flow by combining the actual furnace condition of the blast furnace.
Some scholars extract the edges of the gas flow infrared image by a least square median ellipse fitting method, and provide a method for identifying the central gas flow distribution mode based on an infrared-sensitive imaging image. The number of infrared images adopted by the technology is small, and the distribution characteristics of the gas flow cannot be completely reflected.
Some scholars extract the central features of the gas flow in the infrared image by performing threshold segmentation processing on the infrared image, and then research the distribution state of the gas flow center by performing digital processing on the central features of the gas flow. Although the technology adopts a large number of infrared images, only one central point of each frame of infrared image is extracted, the characteristics of the whole infrared image cannot be represented, and the information quantity is very small.
Disclosure of Invention
Therefore, the invention aims to provide a method for processing gas flow distribution based on the characteristic values of an infrared image pixel matrix.
Specifically, the method for processing the distribution of the gas flow based on the characteristic values of the infrared image pixel matrix comprises the following steps of:
step1, data acquisition and processing: acquiring infrared video production data on line, and acquiring infrared image data of 24 frames per second through image extraction software;
step2, preprocessing infrared image data, wherein the specific process is realized by the following steps:
step 2A, image superposition processing: carrying out batch superposition processing on the infrared image data obtained in the step1 to obtain 3600 frames of infrared images of the gas flow per hour;
step 2B, dead angle removing fuzzification treatment: performing dead angle removing and blurring processing on the infrared image obtained by image superposition processing to prevent the influence of dead angles on characteristic value extraction;
and step 2C, filtering: carrying out mean value and median filtering processing on the infrared image subjected to dead angle blurring removal processing, and filtering noise and pulse interference;
step 2D, pixel matrix square arraying treatment: complementing a 288 × 352 infrared image pixel matrix obtained after filtering into a 352 × 352 square matrix;
step3, solving the maximum eigenvalue of the pixel matrix of each frame of image and the corresponding eigenvector: solving the maximum eigenvalue of the pixel square matrix obtained in the step 2D and the corresponding eigenvector;
and 4, classifying the images by using the sizes of the characteristic values: sorting the maximum eigenvalues found: lambda1|≤|λ2|≤…≤|λk|≤|λk+1|…≤|λk+n|≤…≤|λMAnd l, wherein M is 720 × 3600, the processed image is 720 hours, 3600 frames of gas flow infrared images are counted in each hour, the gas flow infrared images are classified according to the interval range of the characteristic values, different classes represent different types of infrared images, and the infrared images are classified by using the characteristic value size of the infrared image pixel matrix.
Further, in the method, the coal gas flow infrared images are classified according to the interval range of the characteristic values, and the specific classification process is realized through the following steps:
step 4A, rough classification: sequencing the 720-hour 3600 frames of gas flow infrared images per hour into lambda according to the magnitude of the characteristic value1|≤|λ2|≤…≤|λk|≤|λk+1|…≤|λk+n|≤…≤|λ720×3600L, determining the total interval of classification [ | λ |)1|,|λ720×3600|]，
Dividing the total interval [ | lambda [)1|，|λ720×3600|]＝[|λ1|,|λk|)∪[|λk|,|λk+n|)∪…∪[|λk+n+m|，|λ720×3600|]The characteristic value is in the interval [ | lambda1|,|λkI) range is the first type, the characteristic value is in the interval [ | Lambdak|,|λk+nI) the corresponding image in the range is classified into a second class and is classified into N classes by analogy, namely the coal gas flow infrared image is classified according to the interval range of the characteristic value,
step 4B, judging the criterion: calculating the included angle of the eigenvector corresponding to the endpoint value of each interval, and recording the eigenvalue lambdajCorresponding feature vector is
Figure BDA0001441589370000031
For the first class [ | λ1|,|λk|)，λ1Corresponding feature vector is
Figure BDA0001441589370000032
λkCorresponding feature vector is
Figure BDA0001441589370000033
Calculating the absolute value | cos theta of cosine of the included angle of the vector1L is as follows:
Figure BDA0001441589370000034
calculating the second class to the Nth class according to the same method;
step 4C, refining and classifying: and adjusting the endpoint value of each interval to enable the cosine absolute value of an included angle between the corresponding feature vectors of the left and right endpoints of the interval to be close to 1.
The invention has the beneficial effects that: the method for processing the coal gas flow distribution based on the characteristic values of the infrared image pixel matrix classifies the large-sample blast furnace coal gas flow infrared images, finds out the dynamic change characteristic of the coal gas flow distribution in the production process of the blast furnace, provides help for further researching the relation between the coal gas flow change and the coal gas utilization rate, and further realizes the real-time monitoring of the coal gas utilization rate.
Compared with the prior art, the method for classifying the images by processing the acquired infrared images, solving the characteristic values and the corresponding characteristic vectors of the image pixel matrix and utilizing the size of the characteristic values of the image pixels has the greatest advantages of large data processing amount and extraction of all information of the infrared images.
Drawings
Fig. 1 is a flowchart of image processing in the embodiment of the present invention.
Fig. 2 is a flow chart of feature value size classification in an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention is provided with reference to the accompanying drawings:
the invention dynamically tracks the change condition of the gas flow in the blast furnace, which comprises the following steps: (1) data acquisition and processing; (2) processing an infrared image; (3) solving the eigenvalue and the eigenvector of the pixel matrix of each frame of image; (4) and classifying the infrared images by using the size of the characteristic value.
The steps are specifically described as follows:
(1) data acquisition and processing
The data adopted by the invention is mainly a large amount of infrared video data. The method comprises the steps of utilizing software to decode and convert infrared video data acquired online to obtain an infrared image (three-dimensional matrix) with 24 frames per second, and utilizing an image processing tool to convert the infrared image with 24 frames per second to a gray image with a two-dimensional matrix.
(2) Infrared image processing
Processing an original infrared image, wherein the processing flow is shown in fig. 1, and the specific implementation steps are as follows:
(2.1) image superimposition processing
And superposing 24 frames of images in the same second to obtain an infrared image of each second so as to meet the sampling period of data measured by the cross temperature thermocouple and provide a reliable basis for the subsequent research of the relationship between the gas flow change and the corresponding gas utilization rate.
(2.2) dead-angle removal fuzzification processing
And performing dead angle removing and fuzzification processing on non-charge level information such as '2013-12-1506: 00: 00' and 'channel three' in the original infrared image, so that the image information is more accurate.
(2.3) image Filter processing
Because the blast furnace ironmaking process is a highly complex process, the infrared image is easy to generate noise and pulse interference, is not beneficial to extracting image characteristics, and needs to be filtered. And filtering the superposed images to filter the influence of noise and pulse on the images. The invention mainly adopts mean filtering and median filtering to process the image together.
Step 1: and selecting a collected frame of infrared image g (x, y).
Step 2: filtering the infrared image g (x, y) by mean value and storing the filtered infrared image g (x, y) in the image g1In (x, y), namely:
Figure BDA0001441589370000051
step 3: average filtered image g1(x, y) median filtering is performed and stored in the image f (x, y).
(2.4) Pixel matrix squaring
The eigenvalue and eigenvector of the matrix can be solved by the knowledge of linear algebra, and the matrix is required to be a square matrix. Because the main characteristic of the blast furnace infrared image is positioned in the middle of the image, the characteristic of the supplemental image edge has little influence on the overall characteristic of the image. Therefore, the infrared image pixel matrix (288 × 352) obtained after the filtering process is converted into a square matrix (352 × 352), wherein 64 lines of images are supplemented from the 289 th line by taking the 288 th line of images as a standard so that the original pixel matrix is converted into the square matrix.
(3) Solving eigenvalue and eigenvector of pixel matrix of each frame of image
And (3) calculating the pixel matrix eigenvalue and the corresponding eigenvector obtained in the step (2.4) by using an image processing tool, and obtaining the eigenvalue and the eigenvector of the pixel matrix of each frame of image.
(4) Classification by eigenvalue magnitude
The characteristic values of the gas flow infrared images in 720 hours (3600 frames per hour) are sequenced: lambda1|≤|λ2|≤…≤|λk|≤|λk+1|…≤|λk+n|≤…≤|λ720×3600Classifying the gas flow infrared images according to the interval range where the characteristic value is located, wherein the specific process is realized through the following substeps:
(4.1) rough Classification (quartering)
a) And (4) average classification: sorting | lambda | of 720-hour (3600 frames per hour) gas flow infrared images according to magnitude of eigenvalues1|≤|λ2|≤…≤|λk|≤|λk+1|…≤|λk+n|≤…≤|λ720×3600L. Firstly, the characteristic value interval [ | lambda ] is divided1|,|λ720×3600|]Is divided into four equal parts
Figure BDA0001441589370000052
The four feature intervals correspond to four types of images G1, G2, G3, G4. (i.e. dividing the image into four categories according to the visual perception of the infrared image, namely 'brightest, brighter, darker and darkest')
b) Judgment criterion (determination of rationality of image classification): calculating the absolute value of the cosine of the included angle of the eigenvector corresponding to the endpoint value of each interval,
the first type image G1 has characteristic value interval of
Figure BDA0001441589370000061
Cosine of included angle
Figure BDA0001441589370000062
The second type of image G2 has characteristic value interval of
Figure BDA0001441589370000063
Cosine of included angle
Figure BDA0001441589370000064
The third type of image G3 has characteristic value interval of
Figure BDA0001441589370000065
Cosine of included angle
Figure BDA0001441589370000066
The fourth type of image G4 has characteristic value interval of
Figure BDA0001441589370000067
Cosine of included angle
Figure BDA0001441589370000068
When 0.95. ltoreq. cos θNWhen | ≦ 1(N ═ 1,2,3,4), it is stated that the feature vectors corresponding to the feature values of all the images of the nth class are almost parallel, and it is determined that all the images of this class have higher similarity, i.e., the classification effect is better; if | cos θNIf the value of the absolute is less than 0.95, the similarity of the Nth type of image is poor, and the endpoint value of the characteristic interval of the Nth type of image needs to be adjusted to reach the judgment standard.
(4.2) refining classification: the step of refining classification is described by taking the first class G1 as an example:
a) first class | cos θ1If < 0.95, the similarity of the first type of image is not high. Corresponding interval of the first class image characteristic value
Figure BDA0001441589370000069
Bisector into
Figure BDA00014415893700000610
And
Figure BDA00014415893700000611
that is, the first type image G1 is divided into two types G11 and G12, and G11 corresponds to a characteristic value interval of G11
Figure BDA00014415893700000612
G12 has a characteristic value interval of
Figure BDA00014415893700000613
b) Judging the reasonability of classification again: calculating the absolute value of the cosine of the included angle of the feature vectors corresponding to the endpoint values of the corresponding intervals of the new classes G11 and G12, and if the absolute value meets the judgment standard | cos theta [ ]11|≥0.95，|cosθ12If | > 0.95, ending the classification; if not, the first class G1 interval is divided
Figure BDA0001441589370000071
Carrying out K anew as in a)1Is divided equally (K)13,4, …, n), the classification is reasonableness judged until the criterion is satisfied
Figure BDA0001441589370000072
The other three types of G2, G3 and G4 have the same refining classification method.
Fig. 2 is a flow chart of classification with eigenvalue intervals, with the input variables: squaring the processed image; calculating the maximum characteristic value and the corresponding characteristic vector, roughly classifying (quartering) the characteristic value interval, and then refining and classifying according to the mode; the output variables are: classification section and class P (P ═ K)1+K2+K3+K4). The different classes represent different types of infrared images, and the infrared images are classified by using the characteristic value of the infrared image pixel matrix.
Taking 720-hour infrared image as a sample, and obtaining a classification interval and corresponding | cos theta according to the algorithmPThe values are shown in Table 1.
TABLE 1
Figure BDA0001441589370000073
(5) Example (take 1 hour 3600 frame image as an example)
Features determined in accordance with the inventionClassifying the blast furnace infrared images in a certain hour in an eigenvalue interval (see table 1), namely solving 3600 maximum eigenvalues of a pixel matrix of 3600 frames of images and judging an interval I to which the eigenvalues belongiThereby determining the category to which the image belongs. Table 2 shows the ratio of the number of each type of image to the total number of 3600 frames in 1 hour
Figure BDA0001441589370000074
TABLE 2
Figure BDA0001441589370000075
The larger the number of data samples (gas flow infrared images) used by the invention is, the more blast furnace gas flow information is contained, and the more accurate the judgment of the blast furnace condition is. If the video data of the gas flow of a certain blast furnace for 1 year (particularly the gas flow information under abnormal furnace conditions) is collected, all possible conditions of the gas flow distribution in the production process of the blast furnace can be determined by using the method, and accurate data guidance is provided for furnace condition diagnosis and prediction.
The method provided by the invention not only can classify and identify a large amount of gas flow infrared image data, but also can predict the furnace condition of the blast furnace by utilizing the infrared image within a specific period of time.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (2)

1. A method for processing gas flow distribution based on eigenvalues of an infrared image pixel matrix is characterized by comprising the following steps:
step1, data acquisition and processing: acquiring infrared video production data on line, and acquiring infrared image data of 24 frames per second through image extraction software;
step2, preprocessing infrared image data, wherein the specific process is realized by the following steps:
step 2A, image superposition processing: carrying out batch superposition processing on the infrared image data obtained in the step1 to obtain 3600 frames of infrared images of the gas flow per hour;
step 2B, dead angle removing fuzzification treatment: performing dead angle removing and blurring processing on the infrared image obtained by image superposition processing to prevent the influence of dead angles on characteristic value extraction;
and step 2C, filtering: carrying out mean value and median filtering processing on the infrared image subjected to dead angle blurring removal processing, and filtering noise and pulse interference;
step 2D, pixel matrix square arraying treatment: complementing a 288 × 352 infrared image pixel matrix obtained after filtering into a 352 × 352 square matrix;
step3, solving the maximum eigenvalue of the pixel matrix of each frame of image and the corresponding eigenvector: solving the maximum eigenvalue of the pixel square matrix obtained in the step 2D and the corresponding eigenvector;
and 4, classifying the images by using the sizes of the characteristic values: sorting the maximum eigenvalues found: lambda1|≤|λ2|≤…≤|λk|≤|λk+1|≤…≤|λk+n|≤…≤|λMAnd l, wherein M is 720 × 3600, the processed image is 720 hours, 3600 frames of gas flow infrared images are counted in each hour, the gas flow infrared images are classified according to the interval range of the characteristic values, different classes represent different types of infrared images, and the infrared images are classified by using the characteristic value size of the infrared image pixel matrix.
2. The method for processing gas flow distribution based on eigenvalues of infrared image pixel matrix as claimed in claim 1, wherein in said method the infrared images of gas flow are classified according to the interval range of eigenvalues, the concrete classification process is realized by the following steps:
step 4A, rough classification: sequencing the 720-hour 3600 frames of gas flow infrared images per hour into lambda according to the magnitude of the characteristic value1|≤|λ2|≤…≤|λk|≤|λk+1|…≤|λk+n|≤…≤|λ720×3600L, determining the total interval of classification [ | λ |)1|,|λ720×3600|]，
Dividing the total interval [ | lambda [)1|，|λ720×3600|]＝[|λ1|,|λk|)∪[|λk|,|λk+n|)∪…∪[|λk+n+m|，|λ720×3600|]The characteristic value is in the interval [ | lambda1|,|λkI) range is the first type, the characteristic value is in the interval [ | Lambdak|,|λk+nI) the corresponding image in the range is classified into a second class and is classified into N classes by analogy, namely the coal gas flow infrared image is classified according to the interval range of the characteristic value,
step 4B, judging the criterion: calculating the included angle of the eigenvector corresponding to the endpoint value of each interval, and recording the eigenvalue lambdajCorresponding feature vector is
Figure FDA0002593908110000021
For the first class [ | λ1|,|λk|)，λ1Corresponding feature vector is
Figure FDA0002593908110000022
λkCorresponding feature vector is
Figure FDA0002593908110000023
Calculating the absolute value | cos theta of cosine of the included angle of the vector1L is as follows:
Figure FDA0002593908110000024
calculating the second class to the Nth class according to the same method;
step 4C, refining and classifying: and adjusting the endpoint value of each interval to enable the cosine absolute value of an included angle between the corresponding feature vectors of the left and right endpoints of the interval to be more than or equal to 0.95.
CN201710991544.2A 2017-10-23 2017-10-23 Method for processing gas flow distribution based on characteristic values of infrared image pixel matrix Active CN107679581B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710991544.2A CN107679581B (en) 2017-10-23 2017-10-23 Method for processing gas flow distribution based on characteristic values of infrared image pixel matrix

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710991544.2A CN107679581B (en) 2017-10-23 2017-10-23 Method for processing gas flow distribution based on characteristic values of infrared image pixel matrix

Publications (2)

Publication Number Publication Date
CN107679581A CN107679581A (en) 2018-02-09
CN107679581B true CN107679581B (en) 2020-09-01

Family

ID=61141322

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710991544.2A Active CN107679581B (en) 2017-10-23 2017-10-23 Method for processing gas flow distribution based on characteristic values of infrared image pixel matrix

Country Status (1)

Country Link
CN (1) CN107679581B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109852748B (en) * 2019-02-27 2020-05-08 内蒙古科技大学 Method for monitoring development process of gas flow in distribution period of blast furnace and predicting gas utilization rate
CN112270237B (en) * 2020-10-22 2023-08-29 中冶南方工程技术有限公司 Blast furnace gas flow state identification method, system and storage medium
CN114581785B (en) * 2022-05-09 2022-07-22 中国科学院自动化研究所 Multi-source heterogeneous data fusion blast furnace gas flow distribution state identification method and system

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105002321A (en) * 2015-06-16 2015-10-28 内蒙古科技大学 Method for processing coal gas flow center dynamic tracking and monitoring coal gas utilization rate

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105002321A (en) * 2015-06-16 2015-10-28 内蒙古科技大学 Method for processing coal gas flow center dynamic tracking and monitoring coal gas utilization rate

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
"Quick change from throat stopper to blast furnace top fittings:New charging systems for blast furnaces";E.Brzoska etc.;《ResearchGate》;20091231;论文第2-3节 *
"一种基于感红外图像处理的高炉中心煤气流分布模式识别方法";崔桂梅等;《信息与控制》;20141231;第43卷(第1期);论文第110-115页 *
"基于图像处理的中心煤气流分布识别方法";王正友;《计算机仿真》;20130930;第30卷(第9期);论文第357-360页 *
"拟Newton法在高阶矩阵中的应用-求解最大特征值及特征向量";何超等;《计算机工程与应用》;20121231;第48卷(第16期);论文第33-36页 *

Also Published As

Publication number Publication date
CN107679581A (en) 2018-02-09

Similar Documents

Publication Publication Date Title
CN113989279B (en) Plastic film quality detection method based on artificial intelligence and image processing
US20230289979A1 (en) A method for video moving object detection based on relative statistical characteristics of image pixels
CN116205919B (en) Hardware part production quality detection method and system based on artificial intelligence
CN107679581B (en) Method for processing gas flow distribution based on characteristic values of infrared image pixel matrix
CN110766689A (en) Method and device for detecting article image defects based on convolutional neural network
CN107085846A (en) Surface Flaw image-recognizing method
WO2016045215A1 (en) Fingerprint image quality determination method and device
CN114882040B (en) Sewage treatment detection method based on template matching
CN113554629A (en) Strip steel red rust defect detection method based on artificial intelligence
Laga et al. Image-based plant stornata phenotyping
CN109934203A (en) A kind of cost-sensitive increment type face identification method based on comentropy selection
CN112862744B (en) Intelligent detection method for internal defects of capacitor based on ultrasonic image
CN114723704A (en) Textile quality evaluation method based on image processing
CN109934077B (en) Image identification method and electronic equipment
JP7147933B2 (en) Water quality analysis method
CN108073940B (en) Method for detecting 3D target example object in unstructured environment
CN108986077B (en) Flotation froth working condition identification method based on dual-tree complex wavelet domain symbiotic augmentation matrix
CN109598200B (en) Intelligent image identification system and method for molten iron tank number
CN106529441B (en) Depth motion figure Human bodys&#39; response method based on smeared out boundary fragment
CN114494305A (en) On-line detection method and system for particle size of green pellets
CN115082427B (en) Aluminum material calendering process parameter control method based on artificial intelligence
CN113435460A (en) Method for identifying brilliant particle limestone image
CN115294109A (en) Real wood board production defect identification system based on artificial intelligence, and electronic equipment
Zhao et al. A machine learning method for state identification of superheat degree with flame interference
CN102494987A (en) Automatic category rating method for microscopic particles in nodular cast iron

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