CN114936581A - Multi-parameter association mining method based on time sequence data segmentation - Google Patents
Multi-parameter association mining method based on time sequence data segmentation Download PDFInfo
- Publication number
- CN114936581A CN114936581A CN202210621752.4A CN202210621752A CN114936581A CN 114936581 A CN114936581 A CN 114936581A CN 202210621752 A CN202210621752 A CN 202210621752A CN 114936581 A CN114936581 A CN 114936581A
- Authority
- CN
- China
- Prior art keywords
- parameter
- trend
- segmentation
- segments
- parameters
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 48
- 230000011218 segmentation Effects 0.000 title claims abstract description 41
- 238000005065 mining Methods 0.000 title claims abstract description 37
- 238000004364 calculation method Methods 0.000 claims abstract description 21
- 238000013139 quantization Methods 0.000 claims abstract description 8
- 238000010219 correlation analysis Methods 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 6
- 230000001174 ascending effect Effects 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 238000012800 visualization Methods 0.000 claims description 3
- 238000012098 association analyses Methods 0.000 claims description 2
- 238000012545 processing Methods 0.000 abstract description 2
- 230000000694 effects Effects 0.000 description 4
- 239000000284 extract Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2465—Query processing support for facilitating data mining operations in structured databases
-
- 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
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Fuzzy Systems (AREA)
- Signal Processing (AREA)
- Probability & Statistics with Applications (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention relates to a multi-parameter association mining method based on time sequence data segmentation, and belongs to the field of big data processing. The method includes initializing parameters needing correlation analysis, analyzing time series data, extracting trend segmentation features of the data to obtain a quantization result in each segmentation feature, calculating a correlation relation based on the quantization result, and outputting a multi-parameter correlation mining result. Meanwhile, the method also supports expert knowledge to improve the accuracy of the correlation result. The method can improve the calculation performance, and only needs to calculate whether the segmentation trends of all the parameters are related, so that the speed of the mining algorithm is greatly improved; the method is based on the parameter association mining algorithm of the trend segmentation, and the intra-segment trend is quantized and calculated without setting a threshold value, so that the universality of the algorithm is improved; the method can also be used for mining under the condition of less historical samples, and can solve the problem that multi-parameter association mining cannot be carried out on fewer samples.
Description
Technical Field
The invention belongs to the field of big data processing, and particularly relates to a multi-parameter association mining method based on time sequence data segmentation.
Background
In the prior art, correlation analysis is strictly performed on sensor time sequence data according to time periods, and calculation and mining need to be performed on unsegmented data. The prior art closest to the scheme comprises an Apriori method FP-Growth method and the like.
The existing calculation process of multi-parameter association mining is generally as follows:
1. carrying out standardization preprocessing operation on the original data to reduce the influence of data bias, amplitude scaling, linear trend and noise on calculation;
2. setting a correlation threshold value, and performing frequent pattern mining on a multi-parameter time sequence by using sliding window operation;
3. and generating a time sequence correlation result.
The existing multi-parameter association mining method needs to set an association threshold, performs feature extraction on original data by using the operation of a sliding window, and then mines multi-parameter features under the same window, and because the data length of application scenes of the multi-parameter association mining algorithm is too long and the data lengths of different parameters are inconsistent, the calculated amount is greatly increased during feature extraction, and the threshold values of different scenes cannot be unified, so that the universality of the algorithm is reduced.
According to the method and the device, trend segmentation is carried out on the original data, sliding window operation is not needed, the calculation amount can be greatly reduced, the setting of an associated threshold value is avoided through quantification and calculation of the trend, and the universality of the algorithm is improved.
Disclosure of Invention
Technical problem to be solved
The invention aims to solve the technical problem of how to provide a multi-parameter association mining method based on time sequence data segmentation so as to solve the problems that the conventional multi-parameter association mining algorithm is large in calculation amount and cannot uniformly reduce the universality of the algorithm for the threshold values of different scenes.
(II) technical scheme
In order to solve the technical problem, the invention provides a multi-parameter association mining method based on time series data segmentation, which comprises the following steps:
aiming at the parameters obtained by the sensor, a patch is extracted in a down-sampling mode by setting a fixed step length to be used as a local data point, and the parameters D obtained by the sensor i (t) defining a patch set P i ＝{P i,1 ,P i,2 ,…,P i,n }，P i,l Representing a time sequence data set in one patch, wherein n represents the number of the patches, calculating a characteristic value of each patch, and extracting a trend turning point after calculating the characteristic value of each patch;
obtaining a plurality of segments according to the turning points of the trend, and recording as: s i ＝{S i,1 ,S i,2 ,…,S i ,K i In which K is i As a parameter D i Number of segments of (t), by S i,k To represent features within each segment, where S i,k ＝{Loc i,k ,Trend i,k ,Fit i,k }; by quantized S i,k Extracting relevant characteristic indexes of the data in the segments: loc i,k ,Trend i,k And Fit i,k ；
Loc i,k Representing a parameter D i (t) a start-stop position and corresponding offset for the kth segmented feature;
Trend i,k ＝{＜flag,num>indicating the current segmentThe method comprises the steps of marking a whole trend, wherein flag represents the trend of a current section, and num is used for representing the number of patches corresponding to the flag;
Fit i,k representing a parameter D i (t) kth segmentation feature Trend i,k The flag in the flag is marked as a result of linear fitting of ascending, descending and horizontal segments;
based on the segmentation result, the incidence relation and incidence degree between the parameters are described through an incidence matrix, wherein the incidence matrix is R ═ R (R ═ R) ij ) N×N N is the number of parameters participating in the correlation analysis; r is calculated according to the following formula ij ：
Wherein S i ,S j Representing the identity of parameter i and parameter j, F (S) i ,S j ) For calculating the degree of association of two parameters, which is defined as follows:
where K denotes the maximum number of characteristic segments for participating in the relevance calculation, i.e., K ═ max (K) i ,K j )；K i Number of segments, K, of parameter i j The number of segments for parameter j;
for representing whether two parameters are synchronized according to the trend feature statistics of the segmentation result, the definition is as follows:
wherein the content of the first and second substances,
k for representing ith parameter 1 Kth of segment and jth parameter 2 Whether the segments are both ascending and descending, Λ is the intersection or the AND, which is defined as follows:
where Φ represents an empty set.
Further, for each P i,l Performing median calculation (p) i,l ) The median value serves as a characteristic value for each patch.
Further, each parameter has a plurality of history data, each of which is time series data.
Further, Loc i,k ＝{start,end,Δ start ,Δ end Where start denotes start time, end denotes cut-off time, Δ start Denotes the offset, Δ, of start end Indicating the amount of end offset.
Further, flag includes rise, fall, level, and oscillation.
Further, Fit i,k Line represents that the fitting type is linear, slope represents the slope of linear fitting, bias represents the intercept of linear fitting, and Fitter error represents the fitting error of linear fitting.
Further, the air conditioner is provided with a fan,
when calculated, the calculation is considered to occur when the calculation shows that the calculation is performed by increasing one by oneThe association synchronization is performed.
Further, the air conditioner is provided with a fan,
the corresponding time window comes from the Loc feature of the segmented result.
Further, the method further comprises: and after obtaining the incidence matrix R, displaying the incidence relation and the incidence degree between the parameters by utilizing a visualization technology.
Furthermore, the method includes initializing parameters needing correlation analysis, analyzing time series data, extracting trend segmentation features of the data to obtain a quantization result in each segmentation feature, calculating a correlation relation based on the quantization result, and outputting a multi-time series data correlation mining result.
(III) advantageous effects
The invention provides a multi-parameter association mining method based on time sequence data segmentation, which has the technical effects that:
1. the method can improve the calculation performance, and only needs to calculate whether the segmentation trends of all the parameters are related, so that the speed of the mining algorithm is greatly improved;
2. the method is based on the parameter association mining algorithm of the trend segmentation, and the intra-segment trend is quantized and calculated without setting a threshold value, so that the universality of the algorithm is improved;
3. the method can also be used for mining under the condition of less historical samples, and can solve the problem that multi-parameter association mining cannot be carried out on fewer samples.
Drawings
FIG. 1 is a schematic diagram of discretely extracting Patch points and trend feature points;
FIG. 2 is a flowchart of a multi-parameter association calculation.
Detailed Description
In order to make the objects, contents and advantages of the present invention clearer, the following detailed description of the embodiments of the present invention will be made in conjunction with the accompanying drawings and examples.
The invention carries out multi-parameter association relation mining based on the segmentation result of time sequence data, 1) carries out trend segmentation on the time sequence data; 2) the trend of the data is quantified and calculated; 3) and extracting and fusing the parameter association relation.
The technical innovation points of the method are mainly divided into three points as follows:
1. segmenting time series data by utilizing down-sampling patch operation and trend turning point extraction
The trend segmentation module in the method improves a trend turning point algorithm, and directly extracts the trend turning points from the noise-dense data and the frequent jump data, so that a large number of dense trend turning points appear, and the data segmentation effect is greatly influenced. Therefore, the method extracts the trend turning point on the basis of the way of discretely taking the patch on the original data, as shown in fig. 1.
Aiming at the parameters obtained by the sensor, a patch is extracted in a down-sampling mode by setting a fixed step length to be used as a local data point, and the parameters D obtained by the sensor i (t) defining a patch set P i ＝{P i,1 ,P i,2 ,…,P i,n }，P i,l Represents a time series data set in one patch, where n represents the number of patches, for each P i,l Performing median calculation mean (p) i,l ) The method is used for representing the characteristic value of each patch, and trend turning point extraction is carried out after the characteristic value of each patch is calculated. The trend turning point can be extracted by adopting a general method.
A parameter has a plurality of historical data, each of which is time series data.
2. Quantizing and computing data segments
From the trend turning points, several segments can be obtained, denoted as:
wherein K i As a parameter D i Number of segments of (t), here by S i,k To represent features within each segment, where S i,k ＝{Loc i,k ,Trend i,k ,Fit i,k }. By quantized S i,k Can be used for data in the segmentExtracting relevant characteristic indexes, wherein the following three main indexes are mainly extracted: loc i,k ,Trend i,k And Fit i,k 。
Loc i,k Representing a parameter D i (t) the start-stop position and corresponding offset of the kth segmented feature, noted as: loc i,k ＝{start,end,Δ start ,Δ end Where start denotes start time, end denotes cut-off time, Δ start Represents the offset of start, Δ end Indicating the amount of end offset.
Trend i,k ＝{＜flag,num>Indicates the global trend flag of the current segment, wherein flag indicates the trend of the current segment, such as: UP (UP), DOWN (DOWN), horizontal (horizontal), and ringing (vibrant), and num is used to indicate the number of taps corresponding to flag.
Fit i,k Representing the parameter D i (t) kth segmentation feature Trend i,k The flag in inner flag is marked as the result of linear fitting of UP (UP), DOWN (DOWN) and horizontal (horizontal) segments. E.g. Fit i,k Line denotes the type of fit is linear, slope:1.5 denotes the slope of the linear fit is 1.5, bias: 3 denotes the intercept of the linear fit is-3, and Fitterror: 0.000001 denotes the fit error of the linear fit is 0.000001.
3. Extracting and fusing parameter association relation
Based on the segmentation result of the previous step, the association relationship and the association degree between the parameters can be described by an association matrix, wherein the association matrix is R ═ (R ═ R ij ) N×N And N is the number of parameters participating in association analysis. R is calculated according to the following formula ij ：
Wherein S i ,S j Representing the identity of parameter i and parameter j, F (S) i ,S j ) For calculating the degree of association of two parameters, which is defined as follows:
where K denotes the maximum number of characteristic segments for participating in the relevance calculation, i.e., K ═ max (K) i ,K j )；K i Number of segments, K, of parameter i j The number of segments for parameter j;
for representing whether two parameters are synchronized according to the trend feature statistics of the segmentation result, the definition is as follows:
wherein the content of the first and second substances,
k for representing ith parameter 1 Kth of segment and jth parameter 2 Whether a segment is both UP (UP) and DOWN (DOWN), and when it appears as one UP and one DOWN, it is also considered that an associated sync has occurred, a is an intersection or an and, which is defined as follows:
where Φ represents an empty set.
4. After the incidence matrix R is obtained, the incidence relation and the incidence degree between the parameters can be displayed by utilizing a visualization technology.
To sum up, the calculation flow of the method is as shown in fig. 2, firstly, parameters to be subjected to correlation analysis are initialized, time series data are analyzed, then, trend segmentation feature extraction is performed on the data, a quantization result in each segmentation feature is obtained, correlation is calculated based on the quantization result, and finally, a multi-time series data correlation mining result is output. Meanwhile, the method also supports expert knowledge to improve the accuracy of the correlation result.
The invention has the technical effects that:
1. the method can improve the calculation performance, and only needs to calculate whether the segmentation trends of each time series data are related, so that the speed of the mining algorithm is greatly improved;
2. the method is based on the trend segmented parameter association mining algorithm, and the intra-segment trend is quantized and calculated without setting a threshold value, so that the universality of the algorithm is improved;
3. the method can also be used for mining under the condition of less historical samples, and can solve the problem that multi-parameter association mining cannot be carried out on fewer samples.
The above description is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, several modifications and variations can be made without departing from the technical principle of the present invention, and these modifications and variations should also be regarded as the protection scope of the present invention.
Claims (10)
1. A multi-parameter association mining method based on time series data segmentation is characterized by comprising the following steps:
aiming at the parameters obtained by the sensor, a patch is extracted in a down-sampling mode by setting a fixed step length to be used as a local data point, and the parameters D obtained by the sensor i (t) defining a patch set P i ＝{P i,1 ,P i,2 ,…,P i,n }，P i,l Indicating time in a patchThe method comprises the steps of obtaining a sequence data set, wherein n represents the number of the patch, calculating a characteristic value of each patch, and extracting a trend turning point after calculating the characteristic value of each patch;
obtaining a plurality of segments according to the turning points of the trend, and recording as:
wherein K i As a parameter D i Number of segments of (t), by S i,k To represent features within each segment, where S i,k ＝{Loc i,k ,Trend i,k ,Fit i,k }; by quantized S i,k Extracting relevant characteristic indexes of the data in the segments: loc i,k ,Trend i,k And Fit i,k ；
Loc i,k Representing a parameter D i (t) a start-stop position and corresponding offset for the kth segmented feature;
Trend i,k ＝{<flag,num>indicating an integral trend mark of the current segment, wherein the flag indicates the trend of the current segment, and num is used for indicating the number of slots corresponding to the flag;
Fit i,k representing a parameter D i (t) kth segmentation feature Trend i,k The flag in the inner part is marked as a result of linear fitting of ascending, descending and horizontal sections;
based on the segmentation result, the incidence relation and the incidence degree between the parameters are described through an incidence matrix, wherein the incidence matrix is R ═ R (R) ij ) N×N N is the number of parameters participating in the correlation analysis; r is calculated according to the following formula ij ：
Wherein S i ,S j Representing the identity of parameter i and parameter j, F (S) i ,S j ) For calculating the degree of association of two parameters, which is defined as follows:
where K denotes the maximum number of characteristic segments for participating in the relevance calculation, i.e., K ═ max (K) i ,K j )；K i Number of segments, K, of parameter i j The number of segments for parameter j;
for representing whether two parameters are synchronized according to the trend feature statistics of the segmentation result, the definition is as follows:
wherein the content of the first and second substances,
k for representing ith parameter 1 Kth of segment and jth parameter 2 Whether the segments are both ascending and descending, Λ is the intersection or the phase-and, which is defined as follows:
where Φ represents an empty set.
2. Multiple based on time series data segmentation as claimed in claim 1The parameter association mining method is characterized in that each P is subjected to i,l Performing median calculation (p) i,l ) The median value is taken as the characteristic value of each patch.
3. The method of claim 1, wherein each parameter has a plurality of historical data, each historical data being time series data.
4. The method of claim 1, wherein Loc is a multi-parameter associative mining method based on time series data segmentation i,k ＝{start,end,Δ start ,Δ end Where start denotes a start time, end denotes a cut-off time, Δ start Represents the offset of start, Δ end Indicating the amount of end offset.
5. The method of claim 4, wherein flag includes up, down, horizontal, and ringing.
6. The method of claim 5, wherein Fit is a multi-parameter association mining method based on time series data segmentation i,k Line denotes that the fitting type is linear, slope denotes the slope of the linear fitting, bias denotes the intercept of the linear fitting, and Fitter error denotes the fitting error of the linear fitting.
7. The method of claim 6, wherein the mining method comprises mining the data segments in a time-series manner,
in calculation, the correlation synchronization is considered to occur when the correlation synchronization appears to rise and fall.
8. The multi-parameter associative digging based on time series data segmentation of claim 1The digging method is characterized in that,
the corresponding time window comes from the Loc feature of the segmented result.
9. The method of claim 1, wherein the method further comprises: and after obtaining the incidence matrix R, displaying the incidence relation and the incidence degree between the parameters by utilizing a visualization technology.
10. The multi-parameter association mining method based on time series data segmentation of any one of claims 1-9, characterized in that the method first initializes parameters to be subjected to association analysis, analyzes time series data, then performs trend segmentation feature extraction on the data to obtain a quantization result in each segmentation feature, calculates association relation based on the quantization result, and finally outputs multi-time series data association mining results.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210621752.4A CN114936581B (en) | 2022-06-01 | 2022-06-01 | Multi-parameter association mining method based on time sequence data segmentation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210621752.4A CN114936581B (en) | 2022-06-01 | 2022-06-01 | Multi-parameter association mining method based on time sequence data segmentation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114936581A true CN114936581A (en) | 2022-08-23 |
| CN114936581B CN114936581B (en) | 2024-04-26 |
Family
ID=82865691
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210621752.4A Active CN114936581B (en) | 2022-06-01 | 2022-06-01 | Multi-parameter association mining method based on time sequence data segmentation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114936581B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103676645A (en) * | 2013-12-11 | 2014-03-26 | 广东电网公司电力科学研究院 | Mining method for association rules in time series data flows |
| WO2015176565A1 (en) * | 2014-05-22 | 2015-11-26 | 袁志贤 | Method for predicting faults in electrical equipment based on multi-dimension time series |
| CN107562865A (en) * | 2017-08-30 | 2018-01-09 | 哈尔滨工业大学深圳研究生院 | Multivariate time series association rule mining method based on Eclat |
| CN110008253A (en) * | 2019-03-28 | 2019-07-12 | 浙江大学 | The industrial data association rule mining and unusual service condition prediction technique of strategy are generated based on two stages frequent item set |
| CN111046084A (en) * | 2019-12-18 | 2020-04-21 | 重庆大学 | Association rule mining method for multivariate time series monitoring data |
| WO2020177366A1 (en) * | 2019-03-07 | 2020-09-10 | 平安科技（深圳）有限公司 | Data processing method and apparatus based on time sequence data, and computer device |
-
2022
- 2022-06-01 CN CN202210621752.4A patent/CN114936581B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103676645A (en) * | 2013-12-11 | 2014-03-26 | 广东电网公司电力科学研究院 | Mining method for association rules in time series data flows |
| WO2015176565A1 (en) * | 2014-05-22 | 2015-11-26 | 袁志贤 | Method for predicting faults in electrical equipment based on multi-dimension time series |
| CN107562865A (en) * | 2017-08-30 | 2018-01-09 | 哈尔滨工业大学深圳研究生院 | Multivariate time series association rule mining method based on Eclat |
| WO2020177366A1 (en) * | 2019-03-07 | 2020-09-10 | 平安科技（深圳）有限公司 | Data processing method and apparatus based on time sequence data, and computer device |
| CN110008253A (en) * | 2019-03-28 | 2019-07-12 | 浙江大学 | The industrial data association rule mining and unusual service condition prediction technique of strategy are generated based on two stages frequent item set |
| CN111046084A (en) * | 2019-12-18 | 2020-04-21 | 重庆大学 | Association rule mining method for multivariate time series monitoring data |
Non-Patent Citations (2)
| Title |
|---|
| 刘意杨;李俊朋;白洪飞;王智凝;: "基于转折点和趋势段的时间序列趋势特征提取", 计算机应用, no. 1, 10 July 2020 (2020-07-10) * |
| 解初;王建东;韩邦磊;王振;: "基于趋势特征聚类的多元相似时间序列的提取", 科学技术与工程, no. 07, 8 March 2020 (2020-03-08) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114936581B (en) | 2024-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113434357B (en) | Log anomaly detection method and device based on sequence prediction | |
| KR101734829B1 (en) | Voice data recognition method, device and server for distinguishing regional accent | |
| CN111601162B (en) | Video segmentation method and device and computer storage medium | |
| CN112966691A (en) | Multi-scale text detection method and device based on semantic segmentation and electronic equipment | |
| US20220398737A1 (en) | Medical image segmentation method based on u-network | |
| CN111178438A (en) | ResNet 101-based weather type identification method | |
| CN102495887B (en) | Video lens partitioning method based on color matrixes of key regions and application thereof | |
| CN112462261A (en) | Motor abnormality detection method and device, electronic equipment and storage medium | |
| Sasirekha et al. | Enhanced techniques for PDF image segmentation and text extraction | |
| CN111241987B (en) | Multi-target model visual tracking method based on cost-sensitive three-branch decision | |
| CN110473224B (en) | Automatic RSF level set image segmentation method based on KL entropy | |
| CN104637045A (en) | Image pixel labeling method based on super pixel level features | |
| CN114936581A (en) | Multi-parameter association mining method based on time sequence data segmentation | |
| US20070150277A1 (en) | Method and system for segmenting phonemes from voice signals | |
| Law et al. | Energy-efficient data acquisition by adaptive sampling for wireless sensor networks | |
| CN110211150B (en) | Real-time visual target identification method with scale coordination mechanism | |
| CN117349406A (en) | Patent information retrieval system and method based on big data | |
| CN114970630B (en) | Model training method for time sequence data multi-scale segmentation | |
| CN113012158B (en) | Image collaborative segmentation method based on depth dense network and edge distance graph weighting | |
| CN112163580B (en) | Small target detection algorithm based on attention mechanism | |
| CN104850740A (en) | Data flow mode matching method based on dynamic time warping | |
| CN110490297B (en) | Intelligent segmentation method for railway turnout action power curve | |
| CN112200831B (en) | Dynamic template-based dense connection twin neural network target tracking method | |
| CN114691918A (en) | Radar image retrieval method and device based on artificial intelligence and electronic equipment | |
| CN114170164A (en) | Copper bar bending detection method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |