CN109446265B - Complex abnormity identification method and identification system based on workflow - Google Patents
Complex abnormity identification method and identification system based on workflow Download PDFInfo
- Publication number
- CN109446265B CN109446265B CN201811052407.3A CN201811052407A CN109446265B CN 109446265 B CN109446265 B CN 109446265B CN 201811052407 A CN201811052407 A CN 201811052407A CN 109446265 B CN109446265 B CN 109446265B
- Authority
- CN
- China
- Prior art keywords
- key
- mode
- workflow
- time
- features
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/94—Hardware or software architectures specially adapted for image or video understanding
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a complex anomaly identification method and an identification system based on a workflow, wherein the identification method comprises an event stream extraction step, an Nth mode characteristic judgment step and a real-time behavior mode identification step. Compared with the prior art, the complex anomaly identification method and the complex anomaly identification system based on the workflow have universality and can carry out effective anomaly detection on the workflow with quantitative measurable characteristics; meanwhile, when the occurrence of the abnormity is identified, the abnormity type is judged and the specific condition of the abnormity characteristic is output; in addition, for the field of ship track abnormity detection, aiming at tasks of specific starting areas and target areas, the ship track is converted into a state sequence, event detection is carried out, and then a flow abnormity record is extracted, so that the method is favorable for accurately determining the specific time of abnormity and the specific operation task flow, and a new efficient thought is provided for recording and processing abnormal information.
Description
Technical Field
The invention relates to the field of anomaly identification, in particular to a complex anomaly identification method and an identification system based on a work flow.
Background
Anomaly detection is an important aspect in data mining for finding objects in a data set that are significantly different from other data. Anomalies are data that are generated by a completely different mechanism, caused by distinctive, not random deviations in the data set. Anomaly detection has wide application in the fields of finance, medical health, network security, ship operation and the like. In the field of ships, with the development and popularization of geographic information technology in recent years, more and more ships are equipped with positioning devices such as GPS. By collecting and analyzing large-scale ship running track data generated by the GPS equipment, people can be helped to reveal many characteristics such as overwater operation dynamics, ship behavior rules and the like. Through the abnormity detection of the ship track, on one hand, the standardization of the whole ship operation can be promoted, on the other hand, the instant rescue treatment of accidents is facilitated, and meanwhile, evidence can be provided for supervision departments.
A vessel often needs to follow a particular course or a vessel in a different project needs to travel between a particular start and stop area during the course of the voyage. However, due to the interest problem, some ships may choose to run illegally or finish the operation task without following the established flow, and the illegal operation causes the loss of engineering quality and entrusters. Detecting these behaviors is therefore key to ensuring engineering quality and improving engineering efficiency. Currently, these violations are conducted by entrusting a proctoring unit to arrange experienced proctoring personnel according to video monitoring and related data report records, but not only are labor costs large, but also are time-inefficient. And the working quality of the supervisor is not controllable, so that the abnormal sailing behavior of the ship is accurately judged, the overall reputation of a ship company is maintained, the behavior of a ship responsible person is effectively restrained, the legal rights and interests of a project can be protected, the project quality and the project efficiency are improved, and the method has very important significance. At present, a reliable and feasible method is urgently needed for the problem of abnormal navigation of ships. In addition, the types of ship behavior anomalies are multiple, and a supervision unit needs to identify different types of anomalies and then process related personnel, so that the problem is a complex anomaly identification problem, and the anomalies need to be identified and accurately classified.
In recent years, the academic community has higher popularity in abnormal track detection research, and more detection methods are proposed, so that the practical value of the method is highlighted. The applicable field of the similar method is not limited to the abnormal detection of the ship track, and the method is applicable to the workflow with quantitative measurable characteristic quantity. However, the existing method can only treat one-field identification as a two-classification problem, and the final result only judges whether the two-field identification is abnormal, so that the requirement for complex abnormal identification cannot be met. Moreover, due to the problems of privacy, complexity and the like of the track data, the currently feasible ship track abnormity detection method is still relatively deficient.
As disclosed in patent document CN107958346A, the method and apparatus for identifying abnormal behavior reduce the workload of analysis processing by predicting performance index data, thereby further improving the efficiency of identifying abnormal behavior. The technical scheme can only carry out normal or abnormal binary judgment on the performance index variable with single dimensionality, has a limited application range and cannot process the multi-dimensional abnormal detection of a complex task flow.
Also, as disclosed in patent document CN105404895A, an abnormal state recognition method and system includes: acquiring abnormal state data of an individual, wherein the abnormal state data comprises abnormal physiological data and/or abnormal behavior data, matching the abnormal state data with a preset abnormal state model, and determining that the individual is in an abnormal state if the abnormal state data is successfully matched with the abnormal state model. According to the technical scheme, abnormal behaviors are matched and judged, only the abnormality can be specifically found in the identification process, a good effect is achieved for the application in the criminal behavior identification field, and universal application cannot be achieved.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a complex anomaly identification method and an identification system based on a workflow.
The complex anomaly identification method based on the workflow provided by the invention comprises the following steps:
and (3) event stream extraction: inputting a working process, dividing the working process into event streams, using quantitative measurable characteristics as judgment basis of the event streams, identifying key events and key characteristics, and detecting the working process;
an Nth mode feature determination step: judging the Nth mode characteristic in any one or more modes of a historical database, machine learning and mode identification;
the N mode characteristic is a set of key characteristic value ranges of the workflow in the N mode state, and N is a natural number; when N is equal to 0, the nth mode is a normal state of the workflow, and when N > 0, the nth mode is an nth abnormal state.
Preferably, the nth mode characteristic determining step includes the steps of:
a historical event stream identification step: identifying key events and key features as a historical database according to the historical data record of the work;
identifying historical behavior patterns: and counting the historical database to obtain the Nth mode feature and the threshold value of the machine learning algorithm.
Preferably, the complex anomaly identification method based on the workflow provided by the invention comprises the following steps:
a real-time behavior pattern recognition step: in the working process, counting key events and key features in real time and comparing the key events and the key features with the characteristics of the Nth mode in real time; when N is 0, if the key features acquired in real time are the subsets of the zeroth mode features, judging that the mode is normal and outputting a result that the mode is normal; when N is larger than 0, if the key features collected in real time are the subset of the N-th mode features, judging that the N-th mode features are abnormal and outputting the N-th mode abnormality and abnormal key feature data; and if the key features are not any subset of the N-th mode features, judging the key features to be abnormal and outputting unknown abnormal and abnormal key feature data.
Preferably, the event stream extracting step includes:
an input step: and reading the input information of the workflow, and extracting key features of the information of the workflow.
Preferably, the workflow comprises a marine operations workflow;
the key characteristics comprise ship automatic identification system coordinates, namely the relation between the relative position of AIS system coordinates and a preset area, and corresponding time and speed characteristics.
Preferably, the updating method of the history database is as follows:
-the key feature data of the real-time feature acquisition part are transferred to the historical database for updating after the workflow is finished; or
The key characteristic data of the real-time characteristic acquisition part are classified and stored after each workflow is finished, and are transferred to a historical database for updating at the set updating time point.
The invention provides a complex anomaly identification system based on a workflow, which comprises the following modules:
an event stream extraction module: inputting a working process, dividing the working process into event streams, using quantitative measurable characteristics as judgment basis of the event streams, identifying key events and key characteristics, and detecting the working process;
an Nth mode feature determination module: judging the Nth mode characteristic in any one or more modes of a historical database, machine learning and mode identification;
the N mode characteristic is a set of key characteristic value ranges of the workflow in the N mode state, and N is a natural number; when N is equal to 0, the nth mode is a normal state of the workflow, and when N > 0, the nth mode is an nth abnormal state.
Preferably, the nth mode characteristic determining module includes the following sub-modules:
a historical event stream identification submodule: identifying key events and key features as a historical database according to the historical data record of the work;
a historical behavior pattern recognition submodule: and counting the historical database to obtain the Nth mode feature and the threshold value of the machine learning algorithm.
Preferably, the complex anomaly identification system based on workflow provided by the invention comprises the following modules:
a real-time behavior pattern recognition module: in the working process, counting key events and key features in real time and comparing the key events and the key features with the characteristics of the Nth mode in real time; when N is 0, if the key features acquired in real time are the subsets of the zeroth mode features, judging that the mode is normal and outputting a result that the mode is normal; when N is larger than 0, if the key features collected in real time are the subset of the N-th mode features, judging that the N-th mode features are abnormal and outputting the N-th mode abnormality and abnormal key feature data; and if the key features are not any subset of the N-th mode features, judging the key features to be abnormal and outputting unknown abnormal and abnormal key feature data.
Preferably, the event stream extraction module includes:
an input module: and reading the input information of the workflow, and extracting key features of the information of the workflow.
Preferably, the workflow comprises a marine operations workflow;
the key characteristics comprise ship automatic identification system coordinates, namely the relation between the relative position of AIS system coordinates and a preset area, and corresponding time and speed characteristics.
Preferably, the updating method of the history database is as follows:
-the key feature data of the real-time feature acquisition part are transferred to the historical database for updating after the workflow is finished; or
The key characteristic data of the real-time characteristic acquisition part are classified and stored after each workflow is finished, and are transferred to a historical database for updating at the set updating time point.
Compared with the prior art, the invention has the following beneficial effects:
1. the complex anomaly identification method and the complex anomaly identification system based on the workflow have universality and can effectively detect the anomalies of the workflow with quantitative measurable characteristics;
2. the complex anomaly identification method and the complex anomaly identification system based on the workflow can judge the anomaly type and output the specific condition of the anomaly characteristic while identifying the occurrence of the anomaly;
3. for the field of ship track abnormity detection, aiming at tasks of specific starting areas and target areas, the ship track is converted into a state sequence, event detection is carried out, and then a process abnormity record is extracted, so that the method is favorable for accurately determining the specific time of abnormity and the specific operation task process, and a new efficient thought is provided for recording and processing abnormal information.
Drawings
Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:
FIG. 1 is a schematic flow chart of a complex anomaly identification method based on a workflow provided by the present invention;
FIG. 2 is a schematic diagram of an event stream recognition method in the complex anomaly recognition method based on workflow provided by the present invention;
fig. 3 is a schematic diagram of statistical data of the working time of a ship in a certain area in the complex anomaly identification method based on the workflow provided by the invention.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that it would be obvious to those skilled in the art that various changes and modifications can be made without departing from the spirit of the invention. All falling within the scope of the present invention.
The complex anomaly identification method based on the workflow provided by the invention comprises the following steps:
and (3) event stream extraction: inputting a working process, dividing the working process into event streams, using quantitative measurable characteristics as judgment basis of the event streams, identifying key events and key characteristics, and detecting the working process;
an Nth mode feature determination step: judging the Nth mode characteristic in any one or more modes of a historical database, machine learning and mode identification;
the N mode characteristic is a set of key characteristic value ranges of the workflow in the N mode state, and N is a natural number; when N is equal to 0, the nth mode is a normal state of the workflow, and when N > 0, the nth mode is an nth abnormal state.
The nth mode characteristic determining step includes the steps of:
a historical event stream identification step: identifying key events and key features as a historical database according to the historical data record of the work;
identifying historical behavior patterns: and counting the historical database to obtain the Nth mode feature and the threshold value of the machine learning algorithm.
Preferably, the complex anomaly identification method based on the workflow provided by the invention comprises the following steps:
a real-time behavior pattern recognition step: in the working process, counting key events and key features in real time and comparing the key events and the key features with the characteristics of the Nth mode in real time; when N is 0, if the key features acquired in real time are the subsets of the zeroth mode features, judging that the mode is normal and outputting a result that the mode is normal; when N is larger than 0, if the key features collected in real time are the subset of the N-th mode features, judging that the N-th mode features are abnormal and outputting the N-th mode abnormality and abnormal key feature data; and if the key features are not any subset of the N-th mode features, judging the key features to be abnormal and outputting unknown abnormal and abnormal key feature data.
The event stream extraction step includes:
an input step: and reading the input information of the workflow, and extracting key features of the information of the workflow.
The workflow comprises a ship operation workflow; the key characteristics comprise ship automatic identification system coordinates, namely the relation between the relative position of AIS system coordinates and a preset area, and corresponding time and speed characteristics.
The updating mode of the historical database is as follows:
-the key feature data of the real-time feature acquisition part are transferred to the historical database for updating after the workflow is finished; or
The key characteristic data of the real-time characteristic acquisition part are classified and stored after each workflow is finished, and are transferred to a historical database for updating at the set updating time point.
The invention provides a complex anomaly identification system based on a workflow, which comprises the following modules:
an event stream extraction module: inputting a working process, dividing the working process into event streams, using quantitative measurable characteristics as judgment basis of the event streams, identifying key events and key characteristics, and detecting the working process;
an Nth mode feature determination module: judging the Nth mode characteristic in any one or more modes of a historical database, machine learning and mode identification;
the N mode characteristic is a set of key characteristic value ranges of the workflow in the N mode state, and N is a natural number; when N is equal to 0, the nth mode is a normal state of the workflow, and when N > 0, the nth mode is an nth abnormal state.
Specifically, the nth mode characteristic determining module includes the following sub-modules:
a historical event stream identification submodule: identifying key events and key features as a historical database according to the historical data record of the work;
a historical behavior pattern recognition submodule: and counting the historical database to obtain the Nth mode feature and the threshold value of the machine learning algorithm.
More specifically, the complex anomaly identification system based on the workflow provided by the invention comprises the following modules:
a real-time behavior pattern recognition module: in the working process, counting key events and key features in real time and comparing the key events and the key features with the characteristics of the Nth mode in real time; when N is 0, if the key features acquired in real time are the subsets of the zeroth mode features, judging that the mode is normal and outputting a result that the mode is normal; when N is larger than 0, if the key features collected in real time are the subset of the N-th mode features, judging that the N-th mode features are abnormal and outputting the N-th mode abnormality and abnormal key feature data; and if the key features are not any subset of the N-th mode features, judging the key features to be abnormal and outputting unknown abnormal and abnormal key feature data.
The event stream extraction module comprises:
an input module: and reading the input information of the workflow, and extracting key features of the information of the workflow.
The workflow comprises a ship operation workflow;
the key characteristics comprise ship automatic identification system coordinates, namely the relation between the relative position of AIS system coordinates and a preset area, and corresponding time and speed characteristics.
The updating mode of the historical database is as follows:
-the key feature data of the real-time feature acquisition part are transferred to the historical database for updating after the workflow is finished; or
The key characteristic data of the real-time characteristic acquisition part are classified and stored after each workflow is finished, and are transferred to a historical database for updating at the set updating time point.
Further, the technical problem to be solved by the preferred embodiment of the present invention is to provide an event stream identification method for ship trajectory data, and on the basis of the event stream identification method, by applying a machine learning correlation algorithm, abnormal behaviors of a ship are detected and abnormal categories are identified. The core idea of the technical scheme of the invention is as follows: 1) the complete workflow is divided into event streams, the track data are processed and divided according to the relation between the AIS coordinate relative position and the preset area, key events are identified, and the workflow is detected. 2) And extracting key features according to historical data, counting the modes of normal operation of the ship and the threshold value of a machine learning algorithm, and identifying the abnormal condition of the working state of the ship by combining event stream judgment. The event stream identification is illustrated below: as shown in fig. 2, it is assumed that the work flow of the ship needs to enter and exit the a area and the B area once. Then a complete workflow can be divided into event streams: enter the area A, go out of the area A, enter the area B and go out of the area B. The abnormal state in the analysis work needs to be analyzed, key nodes are extracted from continuously obtained ship position coordinate information, and the key nodes are marked as key events. The condition of multiple access can occur in a working area when the ship digs mud. For accurate identification, the time of the last exit from the area a, i.e. the exact time, needs to be found.
Furthermore, the method for detecting ship navigation anomaly in the invention is a general theoretical framework, and takes the example that one complete operation of a ship needs to enter a mud throwing area after the operation in a working area, and finally leaves the mud throwing area to be recorded as completing one task, and some steps of the specific implementation mode of the invention are explained in detail:
the key event and key feature identification algorithm is as follows:
the method comprises the following steps: the first time the ship enters the working area is marked as an event of entering the working area. The time to enter the working area is recorded as tindred；
Step two: marking the event that a ship enters a mud throwing area for the first time as entering the mud throwing area, and recording the time t of entering the mud throwing areaindumpFrom this moment, the last time of exiting the working area is found and recorded as the time t of exiting the working areaexitdred；
Step three: marking the vessel at tindumpThen the mud discharging and throwing area is a mud discharging and throwing area event which is recorded as texitdump。
In the nth mode determining step, through historical data analysis, event stream division and classified statistics are performed on the collected ship working trajectory data based on the collected ship working trajectory data, so as to obtain a mode of a normal working state of the ship, as shown in fig. 3, it can be seen that the distribution of the working time of the ship in the area has strong regularity, and two normal distributions are formed around 150min and 500min respectively. Wherein the distribution taking 150min as the center accounts for more than 80% of the total recorded ship behaviors, namely normal behaviors. Therefore, the threshold value of the stay time of the ship in the normal working state in the final machine learning algorithm in the region is not more than 6 h.
And storing the working records obtained by the real-time anomaly detection into a database, and updating the algorithm threshold value at intervals according to requirements so as to adapt to the change of the actual condition, thereby improving the accuracy and the adaptability of the algorithm. The overall architecture of the algorithm is shown in fig. 1.
Those skilled in the art will appreciate that, in addition to implementing the systems, apparatus, and various modules thereof provided by the present invention in purely computer readable program code, the same procedures can be implemented entirely by logically programming method steps such that the systems, apparatus, and various modules thereof are provided in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Therefore, the system, the device and the modules thereof provided by the present invention can be considered as a hardware component, and the modules included in the system, the device and the modules thereof for implementing various programs can also be considered as structures in the hardware component; modules for performing various functions may also be considered to be both software programs for performing the methods and structures within hardware components.
The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes or modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention. The embodiments and features of the embodiments of the present application may be combined with each other arbitrarily without conflict.
Claims (1)
1. A complex anomaly identification method based on a workflow is characterized by comprising the following steps:
and (3) event stream extraction: inputting a working process, dividing the working process into event streams, using quantitative measurable characteristics as judgment basis of the event streams, identifying key events and key characteristics, and detecting the working process;
an Nth mode feature determination step: judging the Nth mode characteristic in any one or more modes of a historical database, machine learning and mode identification;
the N mode characteristic is a set of key characteristic value ranges of the workflow in the N mode state, and N is a natural number;
when the work flow is normal, the Nth mode is a work flow normal state,
if so, the Nth mode is an Nth abnormal state;
the nth mode characteristic determining step includes the steps of:
a historical event stream identification step: identifying key events and key features as a historical database according to the historical data record of the work;
identifying historical behavior patterns: counting a historical database to obtain the N mode characteristics and a threshold value of a machine learning algorithm;
a real-time behavior pattern recognition step: in the working process, counting key events and key features in real time and comparing the key events and the key features with the characteristics of the Nth mode in real time;
if the key features acquired in real time are the subsets of the zeroth mode features, judging that the key features are normal and outputting a result that the key features are normal;
if the key features collected in real time are the subset of the Nth mode features, judging that the features are abnormal and outputting the Nth mode abnormality and abnormal key feature data; if the key feature is not a subset of any of the N-th mode features, determining an abnormal combinationOutputting unknown anomalies and abnormal key characteristic data;
the event stream extraction step includes:
an input step: reading input information of a workflow, and extracting key features of the information of the workflow;
the workflow comprises a ship operation workflow;
the key characteristics comprise the relation between the relative position of the coordinates of the automatic ship identification system and a preset area, and corresponding time and speed characteristics;
the updating mode of the historical database is as follows:
-the key feature data of the real-time feature acquisition part are transferred to the historical database for updating after the workflow is finished; or
The key characteristic data of the real-time characteristic acquisition part are classified and stored after each work flow is finished, and are transferred to a historical database at the set updating time point to be updated;
the key event and key feature identification steps are as follows:
the method comprises the following steps: marking that the first time the ship enters the working area as an event of entering the working area, and recording the time of entering the working area as tindred；
Step two: marking the event that a ship enters a mud throwing area for the first time as entering the mud throwing area, and recording the time t of entering the mud throwing areaindumpFrom this moment, the last time of exiting the working area is found and recorded as the time t of exiting the working areaexitdred；
Step three: marking the vessel at tindumpThen the mud throwing area is taken as a mud throwing area event, and the mud throwing area time is recorded as texitdump；
And the Nth mode characteristic judging step, namely performing event stream division and classified statistics on the basis of the collected ship working track data through historical data analysis to obtain the mode of the normal working state of the ship.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811052407.3A CN109446265B (en) | 2018-09-10 | 2018-09-10 | Complex abnormity identification method and identification system based on workflow |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811052407.3A CN109446265B (en) | 2018-09-10 | 2018-09-10 | Complex abnormity identification method and identification system based on workflow |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109446265A CN109446265A (en) | 2019-03-08 |
| CN109446265B true CN109446265B (en) | 2021-07-30 |
Family
ID=65531038
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811052407.3A Active CN109446265B (en) | 2018-09-10 | 2018-09-10 | Complex abnormity identification method and identification system based on workflow |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109446265B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103107907A (en) * | 2013-01-04 | 2013-05-15 | 西安交大捷普网络科技有限公司 | Safe responding method based on event flow adding promotion pattern |
| CN104899263A (en) * | 2015-05-22 | 2015-09-09 | 华中师范大学 | Ship trajectory mining, analysis and monitoring method based on specific region |
| CN105653427A (en) * | 2016-03-04 | 2016-06-08 | 上海交通大学 | Log monitoring method based on abnormal behavior detection |
-
2018
- 2018-09-10 CN CN201811052407.3A patent/CN109446265B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103107907A (en) * | 2013-01-04 | 2013-05-15 | 西安交大捷普网络科技有限公司 | Safe responding method based on event flow adding promotion pattern |
| CN104899263A (en) * | 2015-05-22 | 2015-09-09 | 华中师范大学 | Ship trajectory mining, analysis and monitoring method based on specific region |
| CN105653427A (en) * | 2016-03-04 | 2016-06-08 | 上海交通大学 | Log monitoring method based on abnormal behavior detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109446265A (en) | 2019-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109214345B (en) | Method for searching driving track of card-changing vehicle based on similarity comparison | |
| CN106557760A (en) | Monitoring system is filtered in a kind of image frame retrieval based on video identification technology | |
| CN111597962A (en) | Storage material anti-theft alarm method and device and electronic equipment | |
| CN112132041A (en) | Community patrol analysis method and system based on computer vision | |
| CN115546260A (en) | Target identification tracking method and device, electronic equipment and storage medium | |
| CN112800975A (en) | Behavior identification method in security check channel based on image processing | |
| Önal et al. | Object detection for safe working environments using YOLOv4 deep learning model | |
| CN109446265B (en) | Complex abnormity identification method and identification system based on workflow | |
| CN110675521B (en) | Abnormal monitoring and analyzing method, device and system for card punching and storage medium | |
| CN109918901A (en) | The method that real-time detection is attacked based on Cache | |
| Flotzinger et al. | Building inspection toolkit: Unified evaluation and strong baselines for damage recognition | |
| Hwang et al. | Cut and continuous paste towards real-time deep fall detection | |
| Fang et al. | Foul prediction with estimated poses from soccer broadcast video | |
| Salazar et al. | Monitoring approaches for security and safety analysis: application to a load position system | |
| Nain et al. | Authentication control system for the efficient detection of hard-hats using deep learning algorithms | |
| CN114662981A (en) | Pollution source enterprise supervision method based on big data application | |
| Mbuli et al. | Root causes analysis and fault prediction in intelligent transportation systems: coupling unsupervised and supervised learning techniques | |
| He et al. | Prognostics and health management of an automated machining process | |
| CN111882135A (en) | Internet of things equipment intrusion detection method and related device | |
| CN113569637B (en) | Personnel track positioning and supervising system based on Internet of things | |
| CN116311080B (en) | Monitoring image detection method and device | |
| CN111160454B (en) | Quick change signal detection method and device | |
| Islam et al. | Carts: Constraint-based analytics from real-time system monitoring | |
| CN117830686B (en) | Intelligent management and control platform system and method for wearing personal protective equipment on intelligent construction site | |
| Tran12 et al. | and Nhat-Quang Hol, 2 Faculty of Civil Engineering, Ho Chi Minh City University of Technology (HCMUT), 268 Ly Thuong Kiet Street, District 10, Ho Chi Minh City, Vietnam 2 Vietnam National University Ho Chi Minh City, Linh Trung Ward, Thu Duc City, Ho Chi |
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 |