CN112075776A - Intensive cabinet safety control method and system based on artificial intelligence - Google Patents
Intensive cabinet safety control method and system based on artificial intelligence Download PDFInfo
- Publication number
- CN112075776A CN112075776A CN202010956759.2A CN202010956759A CN112075776A CN 112075776 A CN112075776 A CN 112075776A CN 202010956759 A CN202010956759 A CN 202010956759A CN 112075776 A CN112075776 A CN 112075776A
- Authority
- CN
- China
- Prior art keywords
- area
- cabinet
- dense
- camera
- image information
- 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.)
- Withdrawn
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47B—TABLES; DESKS; OFFICE FURNITURE; CABINETS; DRAWERS; GENERAL DETAILS OF FURNITURE
- A47B63/00—Cabinets, racks or shelf units, specially adapted for storing books, documents, forms, or the like
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47B—TABLES; DESKS; OFFICE FURNITURE; CABINETS; DRAWERS; GENERAL DETAILS OF FURNITURE
- A47B53/00—Cabinets or racks having several sections one behind the other
- A47B53/02—Cabinet systems, e.g. consisting of cabinets arranged in a row with means to open or close passages between adjacent cabinets
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47B—TABLES; DESKS; OFFICE FURNITURE; CABINETS; DRAWERS; GENERAL DETAILS OF FURNITURE
- A47B97/00—Furniture or accessories for furniture, not provided for in other groups of this subclass
Abstract
The invention is suitable for the field of artificial intelligence and discloses a safe control method and a safe control system for a dense cabinet based on artificial intelligence. Acquiring a moving track of a person entering a monitoring area through a first camera to obtain first moving track information; the monitoring area comprises a first interested area, and the first interested area is an area where the dense cabinet is located; executing a preset first operation when the first moving track enters the first region of interest; the first operation comprises triggering a second camera to start; acquiring area image information of the first region of interest through the second camera, and feeding back the ground area image information to a control system; the image information of the affiliated area is image information of aisle areas among the dense cabinets; and when the area image information meets the preset conditions, the control system controls to close the cabinet door of the dense cabinet. The operation of automatically closing the dense cabinet is realized.
Description
Technical Field
The invention is suitable for the field of artificial intelligence and discloses a safe control method and a safe control system for a dense cabinet based on artificial intelligence.
Background
At present, the concept of intelligent archives has become popular, and intelligent file cabinets are commonly used in intelligent archives. The intensive cabinet is generally opened and closed by manually operating a control switch or a rotary handle on the intensive cabinet, but files are exposed in the non-constant-temperature and constant-humidity environment due to the fact that the cabinet door is forgotten to be closed, and storage of the files is not facilitated.
Based on the above problems, the current solution is to use a file management system to allow the manager to operate the switch of the cabinet. However, the method needs to sense whether foreign matters exist in the dense cabinet or whether people stay in the dense cabinet, otherwise, potential safety hazards of people and objects are caused.
In the dense cabinet (rack) control system disclosed in the prior art, the aisle condition is sensed through a sensor, and the dense cabinet aisle is sensed by adopting an infrared sensor, an electromagnetic induction sensor and the like, so that the safety of people is ensured.
In practice, the inventors found that the above prior art has the following disadvantages:
the distance and accuracy of the sensors are limited.
Disclosure of Invention
In order to solve the problems, the invention provides a safe control method and a system of a dense cabinet based on artificial intelligence.
In a first aspect, an embodiment of the present invention provides a dense cabinet safety control method based on artificial intelligence, including the following steps:
acquiring a moving track of a person entering a monitoring area through a first camera to obtain first moving track information; the monitoring area comprises a first interested area, and the first interested area is an area where the dense cabinet is located;
executing a preset first operation when the first moving track enters the first region of interest; the first operation comprises triggering a second camera to start;
acquiring regional image information of the first region of interest through the second camera, and feeding back the regional image information to a control system; the image information of the affiliated area is image information of aisle areas among the dense cabinets;
and when the area image information meets the preset conditions, the control system controls to close the cabinet door of the dense cabinet.
Preferably, the executing the preset first operation further includes the steps of:
operating a file instruction according to a preset request to obtain a target file position;
and triggering the cabinet door of the dense cabinet at the target file position to open.
Preferably, the first camera presets a first sampling rate and a second sampling rate, and the second sampling rate is greater than the first sampling rate;
the monitoring region comprises a second region of interest, the second region of interest being different from the first region of interest;
after the person enters the second region of interest, the initial first sampling rate is adjusted to the second sampling rate.
Preferably, the first camera presets a first sampling rate and a second sampling rate, and the second sampling rate is greater than the first sampling rate; the monitoring region includes a second region of interest, the second region of interest being outside the first region of interest.
After the person enters the second region of interest, the initial first sampling rate is adjusted to the second sampling rate.
Preferably, the method further comprises the following steps after the dense cabinet door is closed:
when the first movement track leaves the monitoring area, the first camera is switched from the second sampling rate to the first sampling rate;
the control system turns off the second camera.
Preferably, an infrared sensor is installed in the aisle of the dense cabinet and used for sensing the existence condition of an aisle object and transmitting the sensed information to the control system; and the control system is used for jointly analyzing the closing condition of the dense cabinet by combining the induction information and the area image information.
In a second aspect, another embodiment of the present invention provides a dense cabinet safety control system based on artificial intelligence, including:
the personnel track sensing module is used for acquiring the movement track of personnel entering the monitoring area through the first camera and acquiring first movement track information; the monitoring area comprises a first interested area, and the first interested area is an area where the dense cabinet is located;
the second camera starting module is used for executing preset first operation when the first moving track enters the first region of interest; the first operation comprises triggering a second camera to start;
the image perception module is used for acquiring regional image information of the first region of interest through the second camera and feeding the regional image information back to the control system; the image information of the affiliated area is image information of aisle areas among the dense cabinets;
and the execution module is used for controlling the cabinet door of the dense cabinet to be closed by the control system when the area image information meets the preset condition.
Preferably, the person trajectory sensing module includes:
the first camera is used for carrying out first image acquisition on a monitoring area;
the personnel perception preprocessing module is used for carrying out normalization processing on the first training data set by taking the collected multiple frames of the first images as the first training data set;
the personnel perception encoder is used for performing feature extraction on the first training data set after normalization processing and outputting a first feature map;
the personnel perception decoder is used for sampling the first feature map and outputting a personnel key point heat map;
and the heat accumulation module is used for carrying out accumulation processing on the personnel key point heat map based on a forgetting algorithm to obtain a personnel moving track.
Preferably, the image sensing module includes:
the second camera is used for acquiring image information of the aisle ground area of the dense cabinet;
the semantic segmentation preprocessing module is used for labeling the pixel type of a second training data set by using collected multi-frame image information of the aisle ground area of the dense cabinet as the second training data set, and the labeled data is subjected to one-hot coding processing;
the semantic segmentation coder is used for extracting the characteristics of the second training data set and outputting a second characteristic map;
the semantic segmentation decoder is used for sampling the second feature map and outputting a semantic segmentation map;
and the image perception output module is used for analyzing the semantic segmentation graph and judging whether the cabinet door of the dense cabinet can be closed or not.
Preferably, the operation performed by the execution module further comprises automatically opening a cabinet door.
Preferably, the system further comprises a sensor module for sensing the presence of the aisle object as sensing information to be transmitted to the execution module. The execution module is used for analyzing the closing condition of the dense cabinet together with the information transmitted by the sensor module and the image perception module and executing the automatic closing operation of the dense cabinet.
Compared with the prior art, the implementation mode of the invention has the following beneficial effects:
1. through the linkage of the first camera and the second camera, images are collected and processed, and the switch safety protection of the dense cabinet is realized by combining a system request instruction. The problem of present intensive cabinet of archives' passageway safety protection only judge according to sensor information, the sensor has obvious distance and precision is solved.
2. Due to the fact that a plurality of sensors need to be configured when no dead angle covering is achieved, and the sensors are complex to maintain, repair and replace. The invention works through two cameras and a certain processing unit, and the system is simple and convenient, so that the invention saves consumables to a certain extent.
3. The first camera can switch the sampling rate according to whether personnel exist in the detection area, and the purpose of saving the power consumption of the camera is achieved.
4. The switch of the dense cabinet is controlled by combining the information of the sensor, so that the control system has higher safety.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a method for controlling the safety of a dense cabinet based on artificial intelligence, which is disclosed by the embodiment of the invention;
fig. 2 is a block diagram of a safety control system of a dense cabinet based on artificial intelligence according to an embodiment of the present invention.
Detailed Description
In the following description, for purposes of explanation and not limitation, specific details are set forth, such as particular system structures, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. It will be apparent, however, to one skilled in the art that the present application may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.
It is to be noted that in the following description, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or terminal that comprises the element.
It should be understood that the order of writing each step in this embodiment does not mean the order of execution, and the order of execution of each process should be determined by its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.
In order to explain the technical solution of the present invention, the following is a description of a specific embodiment.
Referring to fig. 1, which is a flowchart of a method for lane safety self-certification based on artificial intelligence and image processing according to an embodiment of the present application, for convenience of description, only a portion related to the embodiment of the present application is shown.
The method comprises the following steps:
step 101: acquiring a moving track of a person entering a monitoring area through a first camera to obtain first moving track information; the monitoring area comprises a first interested area which is the area where the dense cabinet is located.
In this embodiment, the first camera may be deployed on a ceiling or other higher position above the archive area, and the camera is set to a tilt and tilt angle, and the pose is fixed, and the angle of view should cover the whole corresponding monitoring area.
The first camera sets two sampling rates, an initial first sampling rate set to a lower sampling rate, in this embodiment 5 frames/s, and a second sampling rate set to a higher sampling rate, in this embodiment 30 frames/s. These two sampling rates are exemplified in the description of the embodiments that follow. Initially, the first camera keeps the first sampling rate to work, and key points of detection personnel are sensed through the key points. And combining BIM (building information model) geographic information, pre-dividing an interested area of the dense cabinet into a first interested area file interested area on a BIM ground plane as a second interested area, comparing the second interested area file interested area with a personnel moving track projected to the BIM ground, adjusting the sampling rate to be 30 frames/s by a camera after the personnel moving track enters the second interested area, and continuously sensing the personnel key points of the acquired image.
Step 102: executing a preset first operation when the first moving track enters the first region of interest; the first operation includes triggering a second camera to start.
When the first position information detected by the first camera judges that the movement track of the person enters the first region of interest, a calling instruction is sent to the second camera at the corresponding position, the second camera is switched from the sleep mode to the start mode, and the regional image information of the first region of interest is acquired.
In addition, the operation of automatically opening the dense cabinet can be realized by combining a request instruction sent by the management system, namely an instruction for requesting to operate the file, and positioning the position of the dense cabinet to be operated in the file by combining RFID (radio frequency identification) equipment preset on the dense cabinet in the file.
Step 103: acquiring area image information of the first region of interest through a second camera, and feeding back the area image information to the control system; the area image information is the image information of the aisle areas among the dense cabinets.
In this embodiment, the second camera may be disposed at the bottom of each cabinet side of the dense cabinet, i.e., below the file storage area, with a fixed pose and a view angle covering the entire aisle ground area.
And after the second camera is activated by the operation, acquiring the ground image of the passing track of the dense cabinet at the corresponding position to perform semantic segmentation operation. In order to prevent false detection, when the key points sense and detect people and the semantic segmentation detects foreign matters or people, the second camera continues to collect 5 frames of images for processing, and when more than 4 frames of images are processed to obtain the same result, the detection result is judged to be an output result. The control system at least comprises a controller and a communication module, wherein the communication module is used for receiving data information transmitted by the second camera and sending a control command to a corresponding execution unit of the dense cabinet. The execution unit may be a relay switch or a motor, etc.
Step 104: and when the regional image information meets the preset conditions, the control system controls to close the cabinet door of the dense cabinet.
When the fact that the personnel leave the first region of interest is judged according to the personnel moving track information, and the processing result can be judged to be closed after semantic segmentation, at the moment, automatic closing operation of the intensive cabinet is executed, and the second camera is restored to the dormant state; when the person trajectory leaves the second region of interest, the first camera adjusts the sampling rate to return to 5 frames/s.
The method is based on artificial intelligence processing, and compared with the traditional method of using the sensor alone, the method provided by the invention has the advantages of larger monitoring range and higher efficiency. And because a large amount of sensors are not used, materials for constructing the archives are saved.
Preferably, an infrared sensor can be installed in the aisle of the dense cabinet to sense objects, so that the control system can judge whether the cabinet door can be closed or not in cooperation with sensor information and image sensing information.
Based on the same inventive concept as the method embodiment, another embodiment of the invention also provides a dense cabinet safety control system based on artificial intelligence.
Referring to fig. 2, the system includes: personnel trajectory perception module 201, second camera starting module 202, image perception module 203 and execution module 204
The personnel trajectory sensing module 201 is used for acquiring a movement trajectory of a person entering a monitoring area through a first camera and acquiring first movement trajectory information; the monitoring area comprises a first interested area, and the first interested area is an area where the dense cabinet is located;
a second camera starting module 202, configured to execute a preset first operation when the first moving trajectory enters the first region of interest; the first operation comprises triggering a second camera to start;
the image perception module 203 is used for acquiring the regional image information of the first region of interest through the second camera and feeding back the regional image information to the control system; the image information of the affiliated area is image information of aisle areas among the dense cabinets;
and the execution module 204 is configured to control the control system to close the cabinet door of the dense cabinet when the area image information meets a preset condition.
Specifically, the specific working mode of the human trajectory sensing module 201 includes:
at present, there are many Networks and open source projects which can better predict the positions of key points, and the method is a mainstream computer vision task, and the method for detecting the positions of key points used in the embodiment of the present invention is briefly described herein, in which a network is extracted based on the key points of CNN (Convolutional Neural Networks) trainers of an Encoder-Decoder.
The key point of the person is type 1, which is the central point between two feet of the person. The representation in the channel output by the heat map of the personnel key points is the hot spots conforming to Gaussian distribution, when a training set is generally labeled, the hot spots centered on the key points are generated by Gaussian kernel convolution and are labeled, and the hot spots are obtained by training by using a mean square error loss function. And (4) normalizing the label data, so that the output hot spot value range is also located in [ 0-1 ].
The details of the person key point extraction network training are as follows:
1. the first camera 301 is used to collect images of the monitored area of the archive, and the collected multi-frame images are used as a training data set, wherein the images include the situations of existence of people. Images contained in the training data set need to be preprocessed through the personnel perception preprocessing module 302, and the normalization method is adopted as preprocessing, so that better convergence of the model is facilitated. Also, the labels should pass the same pre-treatment.
2. Human keypoint encoder 302 and human keypoint decoder 303 are trained end-to-end through the captured images and heat map label data. The personnel key point encoder 302 performs feature extraction on the image, inputs the image data after normalization processing, and outputs the image data as a feature map; the human key point decoder 303 samples the feature map and outputs a human key point heat map having a size equal to that of the original image, and the key point heat map is output in the form of 1 channel according to the type of the key point because there are 1 types of key points.
3. And training the network by adopting a mean square error loss function.
The heat accumulation module 304 performs accumulation processing on the generated personnel key point heat map based on a forgetting algorithm, and the specific method is as follows:
a) divided into images H of the heat map piled up at the previous moment0The currently obtained detection result HDProcessed heat map H1。
b) For the detection result HDThe heat value should be initialized as high and highest as possibleBase line value B ofstartThe empirical value was 0.75.
c) Thus there is ElementwiseMAXOperator, pair H0、HDAnd a base line value BstartThe method comprises the following steps:
H1＝ElementWiseMAX(H0,HD*Bstart)
d) then, for H1The thermal accumulation attenuation treatment is carried out to make the residual characteristics tend to disappear, and an attenuation coefficient alpha is introduced, and the empirical value is 0.975:
e)H1＝H1*α+HD*(1-α)
thus, the movement track of the person is obtained. The personnel trajectory output module 306 outputs the personnel trajectory to judge the condition of the personnel trajectory entering and exiting the first region of interest and the second region of interest.
The image perception module 202 consists of: a second camera 401, a semantic segmentation encoder 402, a semantic segmentation decoder 403, and an image perception output module 405.
The image sensing module 203 specifically works as follows:
1. adopt second camera 401 to carry out image acquisition to intensive cabinet passageway region to many ground area images are as training data set, and wherein should include personnel, and all kinds of foreign matter existence condition, the foreign matter kind is common conditions such as archives box, water, paper. The pixel classes are labeled by a semantic segmentation pre-processing module 402. The categories are four, respectively human, ground, water, foreign matter and other unrelated items. The foreign matters comprise file boxes, paper and the like. The irrelevant item refers to image information irrelevant to a desired category, such as a dense cabinet appearing in a captured image. The irrelevant item category index is 0, the human category index is 1, the ground category index is 2, the water category index is 3, and the foreign object index is 4. The marking data should be one-hot coded (one-hot coded).
2. The semantic segmentation encoder 403 and semantic segmentation decoder 404 are trained end-to-end through the image and annotation data. The semantic segmentation encoder 403 extracts features and outputs a feature map; the feature map is up-sampled by the semantic segmentation decoder 404, and a semantic segmentation map having a size equal to that of the original image is output.
3. And training the network by adopting a cross entropy loss function.
4. The image sensing output module 405 performs certain post-processing on the output semantic segmentation map. And judging whether the cabinet door of the dense cabinet is allowed to be closed or not according to the pixel value of each pixel which is segmented according to the semantics. An empirical threshold m is set to prevent false detection. When a pixel of a certain pixel value exceeds a threshold value m, it is determined that the category exists. When people, foreign matters, the ground and irrelevant items exist, the cabinet door of the dense cabinet is judged to be not closable; when water, the ground and irrelevant items exist, the cabinet door of the dense cabinet can be closed; when only the ground and the irrelevant items exist, the cabinet door of the dense cabinet can be closed. It should be noted that, when an abnormal pixel position occurs, that is, the pixel position occupying the ground in a general situation exceeds the threshold m, it is also determined that the shutdown is not possible at this time, and abnormal information is sent to the management system.
The execution module 204 receives the results of the personnel trajectory sensing module 201 and the image sensing module 203 to judge the second region of interest of the first region of interest in the monitored region, and intelligently controls and manages the whole archive region through the received instruction so as to guarantee the region safety.
Preferably, the sensor module 205 is provided for sensing dense cabinet aisle objects. The sensor used is an infrared sensor. The infrared sensor used in the present embodiment is widely used and will not be described herein.
When the person trajectory leaves the first region of interest and the corresponding sensor module does not sense an object, the object includes a person or other foreign matter, and the image sensing result determines that the cabinet door of the dense cabinet can be closed, at this time, the execution module 204 closes the cabinet door of the dense cabinet. The intensive cabinet door is cooperatively controlled by the sensing information and the image sensing information of the sensor module, so that the system efficiency is higher, the judgment information is more accurate, and the problem is solved more safely.
It should be noted that: the precedence order of the above embodiments of the present invention is only for description, and does not represent the merits of the embodiments. And specific embodiments thereof have been described above. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. A dense cabinet safety control method based on artificial intelligence is characterized by comprising the following steps:
acquiring a moving track of a person entering a monitoring area through a first camera to obtain first moving track information; the monitoring area comprises a first interested area, and the first interested area is an area where the dense cabinet is located;
executing a preset first operation when the first moving track enters the first region of interest; the first operation comprises triggering a second camera to start;
acquiring area image information of the first region of interest through the second camera, and feeding back the ground area image information to a control system; the image information of the affiliated area is image information of aisle areas among the dense cabinets;
and when the area image information meets the preset conditions, the control system controls to close the cabinet door of the dense cabinet.
2. The artificial intelligence based dense cabinet safety control method according to claim 1, wherein the performing of the preset first operation further comprises the steps of:
operating a file instruction according to a preset request to obtain a target file position;
and triggering the cabinet door of the dense cabinet at the target file position to open.
3. The artificial intelligence based dense cabinet safety control method according to claim 1, wherein the first camera presets a first sampling rate and a second sampling rate, the second sampling rate being greater than the first sampling rate;
the monitoring region comprises a second region of interest, the second region of interest being different from the first region of interest;
after the person enters the second region of interest, the initial first sampling rate is adjusted to the second sampling rate.
4. The artificial intelligence based dense cabinet safety control method according to claim 3, further comprising the following steps after the close of the dense cabinet door:
when the first movement track leaves the monitoring area, the first camera is switched from the second sampling rate to the first sampling rate;
the control system turns off the second camera.
5. The intensive cabinet safety control method based on artificial intelligence as claimed in claim 1, wherein an infrared sensor is installed in the intensive cabinet aisle for sensing the existence of an aisle object and transmitting the sensed information to a control system; and the control system is used for jointly analyzing the closing condition of the dense cabinet by combining the induction information and the area image information.
6. The utility model provides a dense cabinet safety control system based on artificial intelligence which characterized in that includes:
the personnel track sensing module is used for acquiring the movement track of personnel entering the monitoring area through the first camera and acquiring first movement track information; the monitoring area comprises a first interested area, and the first interested area is an area where the dense cabinet is located;
the second camera starting module is used for executing preset first operation when the first moving track enters the first region of interest; the first operation comprises triggering a second camera to start;
the image perception module is used for acquiring regional image information of the first region of interest through the second camera and feeding the regional image information back to the control system; the image information of the affiliated area is image information of aisle areas among the dense cabinets;
and the execution module is used for controlling the cabinet door of the dense cabinet to be closed by the control system when the area image information meets the preset condition.
7. The artificial intelligence based dense cabinet safety control system according to claim 6, wherein the personnel trajectory awareness module comprises:
the first camera is used for carrying out first image acquisition on a monitoring area;
the personnel perception preprocessing module is used for carrying out normalization processing on the first training data set by taking the collected multiple frames of the first images as the first training data set;
the personnel perception encoder is used for performing feature extraction on the first training data set after normalization processing and outputting a first feature map;
the personnel perception decoder is used for sampling the first feature map and outputting a personnel key point heat map;
and the heat accumulation module is used for carrying out accumulation processing on the personnel key point heat map based on a forgetting algorithm to obtain a personnel moving track.
8. The artificial intelligence based dense cabinet safety control system according to claim 6, wherein the image perception module comprises:
the second camera is used for acquiring image information of the aisle ground area of the dense cabinet;
the semantic segmentation preprocessing module is used for labeling the pixel type of a second training data set by using collected multi-frame image information of the aisle ground area of the dense cabinet as the second training data set, and the labeled data is subjected to one-hot coding processing;
the semantic segmentation coder is used for extracting the characteristics of the second training data set and outputting a second characteristic map;
the semantic segmentation decoder is used for sampling the second feature map and outputting a semantic segmentation map;
and the image perception output module is used for analyzing the semantic segmentation graph and judging whether the cabinet door of the dense cabinet can be closed or not.
9. The artificial intelligence based dense cabinet safety control system according to claim 6, wherein the operations performed by the execution module further comprise automatically opening a cabinet door.
10. The artificial intelligence based dense cabinet safety control system according to claim 6, further comprising a sensor module for sensing the presence of an aisle object as sensed information communicated to the executive module. The execution module is used for analyzing the closing condition of the dense cabinet together with the information transmitted by the sensor module and the image perception module and executing the automatic closing operation of the dense cabinet.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010956759.2A CN112075776A (en) | 2020-09-12 | 2020-09-12 | Intensive cabinet safety control method and system based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010956759.2A CN112075776A (en) | 2020-09-12 | 2020-09-12 | Intensive cabinet safety control method and system based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112075776A true CN112075776A (en) | 2020-12-15 |
Family
ID=73737672
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010956759.2A Withdrawn CN112075776A (en) | 2020-09-12 | 2020-09-12 | Intensive cabinet safety control method and system based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112075776A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114332779A (en) * | 2022-03-15 | 2022-04-12 | 云丁网络技术(北京)有限公司 | Method for monitoring target object and related equipment |
| CN117690166A (en) * | 2024-02-02 | 2024-03-12 | 湖北世纪森源电气集团有限公司 | Security monitoring method and system for electric control cabinet |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE3242953A1 (en) * | 1982-11-20 | 1984-05-24 | Bode-Panzer Ag, 3000 Hannover | Movable shelf system, especially lockable drawer system for banks |
| JPH1142127A (en) * | 1997-07-25 | 1999-02-16 | Bunshiyoudou:Kk | Safety picture-monitoring system for use in moving rack device |
| CN201452208U (en) * | 2009-05-25 | 2010-05-12 | 陆琦 | Movable rack with video monitoring device |
| CN103297755A (en) * | 2013-05-13 | 2013-09-11 | 广东欧珀移动通信有限公司 | Video monitoring method and video monitoring terminal |
| CN206349755U (en) * | 2016-12-08 | 2017-07-21 | 北京融安特智能科技股份有限公司 | Compact shelving automatic power-off protection device |
| CN108158234A (en) * | 2017-12-29 | 2018-06-15 | 广州翔特钢艺办公设备有限公司 | Intelligent compact shelf |
| CN108260965A (en) * | 2017-12-29 | 2018-07-10 | 广州翔特钢艺办公设备有限公司 | Intelligent compact shelf |
| CN108521557A (en) * | 2018-04-10 | 2018-09-11 | 陕西工业职业技术学院 | A kind of monitoring method and monitoring system being suitable for large-scale logistics warehouse camera |
| CN110250757A (en) * | 2019-06-19 | 2019-09-20 | 宁波圣达智能科技有限公司 | Intelligent compact shelf and its control method |
-
2020
- 2020-09-12 CN CN202010956759.2A patent/CN112075776A/en not_active Withdrawn
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE3242953A1 (en) * | 1982-11-20 | 1984-05-24 | Bode-Panzer Ag, 3000 Hannover | Movable shelf system, especially lockable drawer system for banks |
| JPH1142127A (en) * | 1997-07-25 | 1999-02-16 | Bunshiyoudou:Kk | Safety picture-monitoring system for use in moving rack device |
| CN201452208U (en) * | 2009-05-25 | 2010-05-12 | 陆琦 | Movable rack with video monitoring device |
| CN103297755A (en) * | 2013-05-13 | 2013-09-11 | 广东欧珀移动通信有限公司 | Video monitoring method and video monitoring terminal |
| CN206349755U (en) * | 2016-12-08 | 2017-07-21 | 北京融安特智能科技股份有限公司 | Compact shelving automatic power-off protection device |
| CN108158234A (en) * | 2017-12-29 | 2018-06-15 | 广州翔特钢艺办公设备有限公司 | Intelligent compact shelf |
| CN108260965A (en) * | 2017-12-29 | 2018-07-10 | 广州翔特钢艺办公设备有限公司 | Intelligent compact shelf |
| CN108521557A (en) * | 2018-04-10 | 2018-09-11 | 陕西工业职业技术学院 | A kind of monitoring method and monitoring system being suitable for large-scale logistics warehouse camera |
| CN110250757A (en) * | 2019-06-19 | 2019-09-20 | 宁波圣达智能科技有限公司 | Intelligent compact shelf and its control method |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114332779A (en) * | 2022-03-15 | 2022-04-12 | 云丁网络技术(北京)有限公司 | Method for monitoring target object and related equipment |
| CN117690166A (en) * | 2024-02-02 | 2024-03-12 | 湖北世纪森源电气集团有限公司 | Security monitoring method and system for electric control cabinet |
| CN117690166B (en) * | 2024-02-02 | 2024-04-16 | 湖北世纪森源电气集团有限公司 | Security monitoring method and system for electric control cabinet |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105700363B (en) | A kind of awakening method and system of smart home device phonetic controller | |
| CN105072740B (en) | A kind of illuminating lamp intelligent control method and system based on Computer Vision | |
| CN112075776A (en) | Intensive cabinet safety control method and system based on artificial intelligence | |
| Lim et al. | iSurveillance: Intelligent framework for multiple events detection in surveillance videos | |
| KR101653278B1 (en) | Face tracking system using colar-based face detection method | |
| US11217076B1 (en) | Camera tampering detection based on audio and video | |
| CN109963046B (en) | Motion detection device and related motion detection method | |
| US11676360B2 (en) | Assisted creation of video rules via scene analysis | |
| US20230196106A1 (en) | Training image classifiers | |
| KR20170052285A (en) | Intelligent camera and System for automatic opening door using Intelligent camera | |
| US11072966B1 (en) | Video analytic enhanced pet door | |
| Dai Duong et al. | An efficient method for vision-based fire detection using SVM classification | |
| CN116246402A (en) | Monitoring method and device | |
| CN112669553A (en) | Unattended system and method for oil and gas station | |
| CN110761680A (en) | Entrance guard anti-trailing device and operation door | |
| CN102024324A (en) | Automatic control system for parking lot barrier | |
| CN108961704A (en) | It reminds vehicle window method for closing and vehicle window is reminded to close device | |
| Sarker et al. | Illegal trash thrower detection based on HOGSVM for a real-time monitoring system | |
| CN105719368A (en) | Personnel detection system and personnel detection method | |
| CN201556298U (en) | Automatic park barrier machine control device | |
| EP4158533A1 (en) | Precipitation removal from video | |
| CN113435371A (en) | Equipment surrounding environment control method, device and equipment based on power internet of things and artificial intelligence | |
| US11783490B2 (en) | Ground plane filtering of video events | |
| EP4095341B1 (en) | Door operating system | |
| US20240020963A1 (en) | Object embedding learning |
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 | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20201215 |
|
| WW01 | Invention patent application withdrawn after publication |