CN110790101A - Elevator trapping false alarm identification method based on big data analysis - Google Patents
Elevator trapping false alarm identification method based on big data analysis Download PDFInfo
- Publication number
- CN110790101A CN110790101A CN201910969449.1A CN201910969449A CN110790101A CN 110790101 A CN110790101 A CN 110790101A CN 201910969449 A CN201910969449 A CN 201910969449A CN 110790101 A CN110790101 A CN 110790101A
- Authority
- CN
- China
- Prior art keywords
- alarm
- elevator
- big data
- trapping
- people
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B5/00—Applications of checking, fault-correcting, or safety devices in elevators
- B66B5/0006—Monitoring devices or performance analysers
- B66B5/0012—Devices monitoring the users of the elevator system
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B5/00—Applications of checking, fault-correcting, or safety devices in elevators
- B66B5/0006—Monitoring devices or performance analysers
- B66B5/0018—Devices monitoring the operating condition of the elevator system
- B66B5/0031—Devices monitoring the operating condition of the elevator system for safety reasons
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B66—HOISTING; LIFTING; HAULING
- B66B—ELEVATORS; ESCALATORS OR MOVING WALKWAYS
- B66B5/00—Applications of checking, fault-correcting, or safety devices in elevators
- B66B5/02—Applications of checking, fault-correcting, or safety devices in elevators responsive to abnormal operating conditions
Landscapes
- Maintenance And Inspection Apparatuses For Elevators (AREA)
- Indicating And Signalling Devices For Elevators (AREA)
Abstract
The invention relates to a method for identifying false alarm of trapping people in elevator based on big data analysis, which combines the characteristics of big data of complex equipment operation and machine analysis theory, provides a trapping people alarming method for the elevator based on big data obtained by a sensor during the actual operation of the elevator, establishes a deep neural network prediction model, acquires signals by an alarm service unit, processes and analyzes the data, judges whether the trapping people is suspected to alarm, if yes, sending alarm information to the people trapping alarm filtering service unit of the Internet of things, carrying out real-time filtering analysis on the alarm by the people trapping alarm filtering service unit based on the Internet of things, judging whether the people trapping alarm filtering service unit is a real person trapping or not according to an analysis result, if the person is really trapped, the trapped alarm information is fed back to the alarm service unit, the alarm service unit informs the maintenance department of the alarm information, and meanwhile, the information storage service unit is started to store the information. The invention can distinguish whether the passenger is really trapped or not in real time, and inform and assist a maintenance unit to timely rescue the trapped passenger in real time.
Description
Technical Field
The invention relates to the field of big data analysis, in particular to a method for judging elevator people trapping and misinformation by utilizing big data analysis.
Background
At present, elevators, particularly vertical elevators, have become standard equipment of numerous residential houses, office buildings and markets, bring convenience to lives of people, and cause frequent accidents due to the fact that maintenance is not in place or manual misoperation and other factors occur. Because the daily cleaning of the vertical elevator, the daily maintenance of the advertisement and the operation and maintenance personnel can cause considerable trapping alarm, the establishment of a perfect and real-time system for trapping elevator people to misinformation is particularly important. At present, the way of alarming to the elevator monitoring sleepers and removing false alarm by the elevator maintenance unit is to establish a set of relatively perfect sleepers alarming logic (combining the opening and closing state of an elevator door, the presence or absence of people in the elevator, the motion state of the elevator and the like), and then to the monitoring video meeting the sleepers alarming logic, the operation and maintenance personnel monitors in real time, and manually distinguishes whether the sleepers alarming is real sleepers, clean-keeping or advertising personnel. The mode of filtering the alarm video of the vertical elevator trapping people is difficult to achieve real-time and high efficiency, meanwhile, operation and maintenance personnel need to watch the alarm monitoring continuously for 24 hours, the labor and the materials are greatly wasted, and misjudgment or misjudgment can be caused by the error of the operation and maintenance personnel, so that dangers of different degrees are caused to elevator taking personnel.
Chinese patent CN 201510232355.8 provides a method and system for detecting people trapping in an elevator. This patent detects whether the elevator is in a trapped state according to a logical combination between a plurality of detection signals for one-step elevators output from an elevator controller, wherein the detection signals include: a man carrying signal, a door lock signal, a fault signal and an overhaul signal; when the trapped state is determined, a corresponding alarm signal is emitted.
The technical scheme has the following defects: the scheme mainly aims at the logic combination among a plurality of detection signals of the one-step elevator to detect whether the elevator is in a trapped state, more signals are needed, when a signal has a fault (such as the fault of an acceleration sensor), the trapped state can be missed, and meanwhile, due to the serial processing of the plurality of signals, the real-time performance of trapped state alarm processing is also greatly influenced; meanwhile, in order to ensure high accuracy, a plurality of expensive sensors (including an acceleration sensor, a PIR infrared sensor and the like) are needed to acquire accurate elevator state signals, so that the cost of the detection system is high.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides an elevator trapping misinformation identification method based on big data analysis, which comprises the following steps of:
step 1: acquiring historical big data of elevator alarm;
step 2: analyzing the big data, extracting the alarm characteristics of the elevator, and taking the alarm characteristics as a training sample;
and step 3: establishing an elevator operation alarm model based on a deep convolutional neural network according to the corresponding relation between the elevator operation characteristic parameters and the elevator operation trapped alarm reasons;
and 4, step 4: in the elevator operation process, elevator operation big data are obtained in real time, alarm characteristic information is extracted, and alarm analysis is performed by utilizing a deep convolutional neural network model.
Preferably, the process and method for acquiring big data of elevator operation comprises the following steps:
step 1: establishing a sensor network for data acquisition;
step 2: collecting current of a permanent magnet synchronous traction machine, vibration parameters of the permanent magnet synchronous traction machine, speed of a lift car, acceleration of the lift car, load of the lift car and noise of a working environment of a lift car box body;
and step 3: and carrying out noise reduction on the acquired information according to a wavelet mode maximum value method, and extracting characteristic information to obtain large elevator running history data which can be used as model training parameters.
Preferably, the method comprises the following steps according to the corresponding relation between the elevator operation alarm and the alarm reason:
step 1: analyzing each characteristic parameter of the elevator historical characteristic big data;
step 2: tracing the root and obtaining the corresponding relation between the alarm and the reason of the alarm according to the prior knowledge; thus, a mapping relation from input X to output Y of the prediction model is established;
and step 3: and coding the alarm reasons, classifying and adding marks according to real trapping, cleaning and billboard replacement, wherein the occurrence probability of each alarm reason is used as the output of a fault prediction model, and the characteristic parameters are used as the input of the fault prediction model.
Preferably, the building and training of the elevator operation alarm model based on the deep convolutional neural network comprises the following steps:
step 1: normalizing the characteristic parameter data to be used as the input of a deep convolution neural network model;
step 2: building a deep convolutional neural network model;
and step 3: based on training data and characteristic parameters as an input vector X of the model, taking the probability of alarming due to trapping as an output Y and actually alarming due to trapping as Y';
and 4, step 4: calculating cross entropy according to the alarm and the alarm reason, and taking the cross entropy as a loss function of the model;
and 5: and training the model by using a batch gradient descent method, and continuously updating the weight and the deviation of the model to minimize a loss function until an elevator alarm model is obtained.
Preferably, the elevator trapping false alarm identification method based on big data analysis comprises the following steps:
(1) the signal acquisition unit acquires information at the elevator door;
(2) the signal analysis module analyzes the state of the elevator door according to the acquired information;
(3) the human body detection module detects whether a person exists in the elevator car;
(4) judging whether the elevator belongs to a suspected alarm condition or not under the condition that people exist, wherein the judgment standard at the moment is that whether the door closing state in the elevator lasts for a period of time or not and the elevator stops running within the lasting period of time or not;
(5) if the suspected alarm condition exists, recording information of a certain time before and after the suspected alarm time point;
(6) and identifying the suspected alarm information by using the trained network, and judging whether the true elevator trapping situation exists.
Further, the signal analysis module is selected from one or more of CPU, ARM, DSP, GPU, FPGA, ASIC and other general processing equipment, and the analysis result of the signal analysis module is the opening and closing state of the elevator door, including opening, closing, opening and closing the door.
Further, the human detection module is a PIR detector.
The invention has the beneficial effects that: according to the conventional running state and operation condition of the elevator car, the condition that people are in the elevator car is divided into true trapping, cleaning and changing of the advertising board, then the network is trained and learned, so that the network has the functions of identifying and distinguishing the three conditions, the possibility of trapping is judged under the condition that the people are in the elevator by combining signal acquisition and PIR detection, and then the information of trapping possibly is further identified through the network, so that an accurate judgment result of trapping of the elevator is obtained, and the trouble of passengers and the waste of resources caused by the misinformation of trapping of the elevator are prevented.
Drawings
FIG. 1 is a diagram of the elevator operational data acquisition hardware configuration of the present invention;
fig. 2 is an overall flowchart.
Fig. 3 is a network model diagram.
Detailed Description
The following detailed description of the embodiments of the present invention is provided in conjunction with the accompanying drawings, and it should be understood that the examples are only illustrative of the technical solutions of the present invention, and are not intended to limit the present invention.
Referring to the attached figures 1-3, the elevator people trapping false alarm identification method based on big data analysis comprises the following steps:
(1) the signal acquisition unit acquires information at the elevator door;
(2) the signal analysis module analyzes the state of the elevator door according to the acquired information;
(3) the human body detection module detects whether a person exists in the elevator car;
(4) judging whether the elevator belongs to a suspected alarm condition or not under the condition that people exist, wherein the judgment standard at the moment is that whether the door closing state in the elevator lasts for a period of time or not and the elevator stops running within the lasting period of time or not;
(5) if the suspected alarm condition exists, recording videos of a certain time before and after the suspected alarm time point;
(6) the method comprises the following steps of identifying a suspected alarm video by using a trained network, and judging whether a real network training process exists, wherein the method comprises the following specific steps:
(A1) collecting the condition that people exist in an elevator car as a training sample;
(A2) classifying the training samples according to real trapping, cleaning and changing billboards, and adding marks;
(A3) utilizing a network to carry out recognition learning;
(A4) and verifying the learned network.
Examples of training and testing methods are as follows:
1. and (4) selecting a sample.
C3D is input as 16 x 240 x 320 image data, where 240 is the image height and 320 is the image width, 16 images in time succession. The input is 16 x 240 x 320 image data, whether training samples or test samples.
2. C3D training process.
A total of 6 ten thousand samples, that is, 6 ten thousand (16 x 240 x 320) image data are trained, each image data is labeled, i.e., set to 0 (true sleepers and forgetting to press the elevator button), set to 1 (clean-up), and set to 2 (change billboard), because the C3D author has pre-trained a network on sport1m (conv3d _ deepnet a _ sport1m _ iter _ 1900000), only the fine t un of the pre-trained network is needed to form a new network. In particular, the num _ output value at the fc8 level in the prototxt file is modified to be 4. Therefore, training samples are collected according to the label example 1:1:1, namely 2 thousands of labels are 0 (real sleepers and forgetting to press elevator buttons), 2 thousands of labels are 1 (cleaning), and 2 thousands of labels are 2 (changing billboards).
And then, using the samples to perform finetune, wherein the size of batch is selected to be 30, namely 30 samples are selected in one iteration, and finally, 6 ten thousand training samples are trained through 2000 iterations. Final SOFTMAX _ LOSS layer output after training
loss＝0.0083。
3. C3D test procedure.
The test process is similar to the training process, the input is (16 × 240 × 320) image data, the number of test samples is 4 ten thousand, each test sample has the label 0 (really sleeps and forgets to press the elevator button), 1 (clean cleaning) and 2 (change the billboard), each test sample is sent into the trained C3D network, the output is 3 values, and the probabilities of inputting 0,1 and 2 are respectively. As shown in the following figures, in this scenario, it can be seen that the probability of judging that the user is really trapped or forgets to press the elevator button is 0.67, the probability of judging that the user is cleaning is 0.11, and the probability of changing the billboard is 0.16. We can judge that the video segment (16 x 240 x 320) is judged to be trapped or forget to press the elevator button.
Similarly, the test is performed on 4 ten thousand test samples, each test sample (16 × 240 × 320) may give a label, and the label is compared with the real label, if the label is consistent, the sample is judged to be correct, otherwise, the sample is judged to be wrong.
The final recognition rate is the ratio of all correctly judged samples in 4 ten thousand samples to the total number of samples. Since the correct sample was determined to be 38080 in actual measurement, the recognition rate was 95.2%.
Claims (6)
1. A method for identifying the trapping and false alarm of an elevator based on big data analysis is characterized by comprising the following steps:
step 1: acquiring historical big data of elevator alarm;
step 2: analyzing the big data, extracting the alarm characteristics of the elevator, and taking the alarm characteristics as a training sample;
and step 3: establishing an elevator operation alarm model based on a deep convolutional neural network according to the corresponding relation between the elevator operation characteristic parameters and the elevator operation trapped alarm reasons;
and 4, step 4: in the elevator operation process, elevator operation big data are obtained in real time, alarm characteristic information is extracted, and alarm analysis is performed by utilizing a deep convolutional neural network model.
2. The elevator fault early warning method based on big data analysis as claimed in claim 1, characterized in that the corresponding relationship between the elevator operation alarm and the reason of the alarm comprises the following steps:
step 1: analyzing each characteristic parameter of the elevator historical characteristic big data;
step 2: tracing the root and obtaining the corresponding relation between the alarm and the reason of the alarm according to the prior knowledge; thus, a mapping relation from input X to output Y of the prediction model is established;
and step 3: and coding the alarm reasons, classifying and adding marks according to real trapping, cleaning and billboard replacement, wherein the occurrence probability of each alarm reason is used as the output of a fault prediction model, and the characteristic parameters are used as the input of the fault prediction model.
3. The elevator manor false alarm method based on big data analysis is characterized by comprising the following steps of:
(1) the signal acquisition unit acquires information at the elevator door;
(2) the signal analysis module analyzes the state of the elevator door according to the acquired information;
(3) the human body detection module detects whether a person exists in the elevator car;
(4) judging whether the elevator belongs to a suspected alarm condition or not under the condition that people exist, wherein the judgment standard at the moment is that whether the door closing state in the elevator lasts for a period of time or not and the elevator stops running within the lasting period of time or not;
(5) if the suspected alarm condition exists, recording signals of a certain time before and after the suspected alarm time point;
(6) and identifying the suspected alarm signal by using the trained network, judging whether the suspected alarm signal is real or not, and judging whether the real elevator trapping condition exists or not.
4. The elevator manor false alarm method based on big data analysis as claimed in claim 1, wherein the method further comprises a process of training the network by using the calibration signal, and the specific steps are as follows:
(A1) collecting the condition that people exist in an elevator car as a training sample;
(A2) classifying the training samples according to real trapping, cleaning and changing billboards, and adding marks;
(A3) carrying out identification analysis by using a network;
(A4) and verifying the analyzed network.
5. The elevator manor false alarm method based on big data analysis as claimed in claim 3, wherein the signal analysis module is selected from one or more of general processing devices of CPU, ARM, DSP, GPU, FPGA, ASIC.
6. The elevator people trapping false alarm method based on big data analysis as claimed in claim 3, wherein the human body detection module is a PIR detector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910969449.1A CN110790101A (en) | 2019-10-12 | 2019-10-12 | Elevator trapping false alarm identification method based on big data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910969449.1A CN110790101A (en) | 2019-10-12 | 2019-10-12 | Elevator trapping false alarm identification method based on big data analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110790101A true CN110790101A (en) | 2020-02-14 |
Family
ID=69439022
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910969449.1A Pending CN110790101A (en) | 2019-10-12 | 2019-10-12 | Elevator trapping false alarm identification method based on big data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110790101A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112347939A (en) * | 2020-11-09 | 2021-02-09 | 浙江新再灵科技股份有限公司 | Ladder-taking non-civilized key identification method based on multi-dimensional features |
| CN112573316A (en) * | 2020-12-08 | 2021-03-30 | 成都睿瞳科技有限责任公司 | Elevator trapping detection method based on computer vision |
| CN113158888A (en) * | 2021-04-19 | 2021-07-23 | 广州咔隆安防科技有限公司 | Elevator abnormal video identification method |
| CN113353755A (en) * | 2021-06-09 | 2021-09-07 | 揭阳市聆讯软件有限公司 | Method for keeping intelligent value of emergency rescue of elevator internet of things |
| CN114455413A (en) * | 2022-01-29 | 2022-05-10 | 广东卓梅尼技术股份有限公司 | Elevator trapping warning method and device and electronic equipment |
| CN114455414A (en) * | 2022-01-29 | 2022-05-10 | 广东卓梅尼技术股份有限公司 | Elevator passenger falling-down warning method and device and electronic equipment |
| CN115724312A (en) * | 2022-05-31 | 2023-03-03 | 海纳云物联科技有限公司 | Method and device for detecting people trapping of elevator car |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006112194A1 (en) * | 2005-04-08 | 2006-10-26 | Mitsubishi Denki Kabushiki Kaisha | Elevator controller |
| CN104803251A (en) * | 2015-05-08 | 2015-07-29 | 上海新时达电气股份有限公司 | Method and system for detecting person trapping in elevator |
| CN104944240A (en) * | 2015-05-19 | 2015-09-30 | 重庆大学 | Elevator equipment state monitoring system based on large data technology |
| CN108178037A (en) * | 2017-12-30 | 2018-06-19 | 武汉大学 | A kind of elevator faults recognition methods based on convolutional neural networks |
| CN108298393A (en) * | 2017-12-20 | 2018-07-20 | 浙江新再灵科技股份有限公司 | Method based on the wrong report of depth network filtering elevator malfunction |
| CN109110608A (en) * | 2018-10-25 | 2019-01-01 | 歌拉瑞电梯股份有限公司 | A kind of elevator faults prediction technique based on big data study |
| CN109292570A (en) * | 2018-10-16 | 2019-02-01 | 宁波欣达（集团）有限公司 | A kind of system and method for elevator technology of Internet of things detection elevator malfunction |
| CN110002303A (en) * | 2018-08-09 | 2019-07-12 | 浙江新再灵科技股份有限公司 | A kind of system and method based on the wrong report of temporal relationship network real time filtering elevator malfunction |
-
2019
- 2019-10-12 CN CN201910969449.1A patent/CN110790101A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006112194A1 (en) * | 2005-04-08 | 2006-10-26 | Mitsubishi Denki Kabushiki Kaisha | Elevator controller |
| CN104803251A (en) * | 2015-05-08 | 2015-07-29 | 上海新时达电气股份有限公司 | Method and system for detecting person trapping in elevator |
| CN104944240A (en) * | 2015-05-19 | 2015-09-30 | 重庆大学 | Elevator equipment state monitoring system based on large data technology |
| CN108298393A (en) * | 2017-12-20 | 2018-07-20 | 浙江新再灵科技股份有限公司 | Method based on the wrong report of depth network filtering elevator malfunction |
| CN108178037A (en) * | 2017-12-30 | 2018-06-19 | 武汉大学 | A kind of elevator faults recognition methods based on convolutional neural networks |
| CN110002303A (en) * | 2018-08-09 | 2019-07-12 | 浙江新再灵科技股份有限公司 | A kind of system and method based on the wrong report of temporal relationship network real time filtering elevator malfunction |
| CN109292570A (en) * | 2018-10-16 | 2019-02-01 | 宁波欣达（集团）有限公司 | A kind of system and method for elevator technology of Internet of things detection elevator malfunction |
| CN109110608A (en) * | 2018-10-25 | 2019-01-01 | 歌拉瑞电梯股份有限公司 | A kind of elevator faults prediction technique based on big data study |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112347939A (en) * | 2020-11-09 | 2021-02-09 | 浙江新再灵科技股份有限公司 | Ladder-taking non-civilized key identification method based on multi-dimensional features |
| CN112573316A (en) * | 2020-12-08 | 2021-03-30 | 成都睿瞳科技有限责任公司 | Elevator trapping detection method based on computer vision |
| CN113158888A (en) * | 2021-04-19 | 2021-07-23 | 广州咔隆安防科技有限公司 | Elevator abnormal video identification method |
| CN113353755A (en) * | 2021-06-09 | 2021-09-07 | 揭阳市聆讯软件有限公司 | Method for keeping intelligent value of emergency rescue of elevator internet of things |
| CN114455413A (en) * | 2022-01-29 | 2022-05-10 | 广东卓梅尼技术股份有限公司 | Elevator trapping warning method and device and electronic equipment |
| CN114455414A (en) * | 2022-01-29 | 2022-05-10 | 广东卓梅尼技术股份有限公司 | Elevator passenger falling-down warning method and device and electronic equipment |
| CN114455414B (en) * | 2022-01-29 | 2024-04-05 | 广东卓梅尼技术股份有限公司 | Elevator passenger overturn alarming method and device and electronic equipment |
| CN114455413B (en) * | 2022-01-29 | 2024-04-05 | 广东卓梅尼技术股份有限公司 | Elevator trapping alarm method and device and electronic equipment |
| CN115724312A (en) * | 2022-05-31 | 2023-03-03 | 海纳云物联科技有限公司 | Method and device for detecting people trapping of elevator car |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110790101A (en) | Elevator trapping false alarm identification method based on big data analysis | |
| CN108569607B (en) | Elevator fault early warning method based on bidirectional gating cyclic neural network | |
| CN110287552B (en) | Motor bearing fault diagnosis method and system based on improved random forest algorithm | |
| CN111362089B (en) | Method and system for identifying entering of electric vehicle into lift car through artificial intelligence | |
| CN109867186B (en) | Elevator trapping detection method and system based on intelligent video analysis technology | |
| CN111680613B (en) | Method for detecting falling behavior of escalator passengers in real time | |
| CN115169505B (en) | Early warning method and early warning system for mechanical fault of moving part of special equipment | |
| CN112193959A (en) | Method and system for detecting abnormal sound of elevator | |
| CN111731960B (en) | Elevator door opening and closing state detection method | |
| CN110589647A (en) | Method for real-time fault detection and prediction of elevator door through monitoring | |
| CN108298393A (en) | Method based on the wrong report of depth network filtering elevator malfunction | |
| CN110008804B (en) | Elevator monitoring key frame obtaining and detecting method based on deep learning | |
| CN109919066B (en) | Method and device for detecting density abnormality of passengers in rail transit carriage | |
| CN108776452B (en) | Special equipment field maintenance monitoring method and system | |
| CN116342895B (en) | Method and system for improving sorting efficiency of renewable resources based on AI (advanced technology attachment) processing | |
| CN113657221A (en) | Power plant equipment state monitoring method based on intelligent sensing technology | |
| CN110002303A (en) | A kind of system and method based on the wrong report of temporal relationship network real time filtering elevator malfunction | |
| CN114436087B (en) | Deep learning-based elevator passenger door-pulling detection method and system | |
| CN112723075B (en) | Method for analyzing elevator vibration influence factors with unbalanced data | |
| CN113156913A (en) | ABS fault diagnosis system and method | |
| CN115724314A (en) | Elevator fault detection early warning system and method | |
| CN114387586A (en) | Driver concentration detection method based on YOLO neural network | |
| CN113247720A (en) | Intelligent elevator control method and system based on video | |
| CN113928947B (en) | Elevator maintenance process detection method and device | |
| CN112347939B (en) | Ladder-taking non-civilized key identification method based on multi-dimensional features |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200214 |
|
| RJ01 | Rejection of invention patent application after publication |