CN108109220A - A kind of classroom work attendance statistics system based on monitoring camera - Google Patents
A kind of classroom work attendance statistics system based on monitoring camera Download PDFInfo
- Publication number
- CN108109220A CN108109220A CN201711483705.3A CN201711483705A CN108109220A CN 108109220 A CN108109220 A CN 108109220A CN 201711483705 A CN201711483705 A CN 201711483705A CN 108109220 A CN108109220 A CN 108109220A
- Authority
- CN
- China
- Prior art keywords
- classroom
- model
- image
- human body
- upper half
- 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
Classifications
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C1/00—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people
- G07C1/10—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people together with the recording, indicating or registering of other data, e.g. of signs of identity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/46—Extracting features or characteristics from the video content, e.g. video fingerprints, representative shots or key frames
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
Landscapes
- Engineering & Computer Science (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 discloses a kind of classroom work attendance statistics systems based on monitoring camera, it is related to work attendance statistics systems technology field;Its work attendance statistics system includes following flow:S101:First, using the method for machine learning, learn a upper half of human body model;S102:The image in the monitor video of classroom is identified using the model in S101, so as to calculate the number of student in the classroom;S103:By being compared with the number that should arrive in each classroom in school education administration system, so as to calculate late number, number of cutting classes and to class rate;The present invention only needs a upper half of human body model succeeded in school in advance, you can applied to all classroom environments, greatly reduces the difficulty that system is implemented;Well adapt to ability;It can be to arriving class rate, late number, number of cutting classes are counted.
Description
Technical field:
The present invention relates to a kind of classroom work attendance statistics systems based on monitoring camera, belong to work attendance statistics systems technology neck
Domain.
Background technology:
For school, especially colleges and universities, it is necessary to periodically be counted to the attendance situation in each classroom, attendance situation bag
It includes:Number, actual arrival number, late number, number of cutting classes etc. should be arrived.Under normal conditions, the administrative staff of school use and manually patrol
The mode looked into carries out, but since classroom is more, it is difficult to patrol each classroom in the same time, can only use selective examination
Mode, it is not comprehensive enough so as to cause the data of statistics.
Currently, the classroom of many schools is assembled with monitoring camera, image recognition technology can be utilized completely, to each
The monitor video image in classroom is identified, and programming count goes out the number in the classroom, so as to quickly and easily to all religions
The attendance situation of room is counted.Chinese patent CN104156729A discloses a kind of classroom demographic method, major function
It is that monitor video is identified, so as to count the number in classroom.The system includes:The back of the body is established under the unmanned state of classroom
Scape model, and establish upper half of human body edge two-value model;Dilation operation is carried out to two-value foreground image, obtains possible human body
Zone of action;The two-value peak point in two-value foreground image is searched for, adds in crown detection set;By upper half of human body edge two-value
Model carries out similarity-rough set with the edge image in the window after slip each time, will be greater than the upper half of human body region of threshold value
Add in human testing set;With reference to crown detection set and human testing set, number system is carried out by the way of cluster analysis
Meter.The main following points of shortcoming of the system:
1st, picture of each classroom under unmanned state is needed, for establishing the background model in classroom.If it counts
Classroom quantity it is more or to be quickly applied to other schools, it will greatly increase the difficulty of system implementation.
2nd, need to set a threshold value, for determining whether upper half of human body region.And the threshold value has very greatly with environment
The problem of relation, environment is different, and threshold value can also change therewith, and the setting of threshold value is one relatively difficult.
3rd, the system has only counted the number in classroom, and late number, number of cutting classes are not counted.
The content of the invention:
In view of the above-mentioned problems, the technical problem to be solved in the present invention is to provide a kind of classroom attendances based on monitoring camera
Statistical system.
A kind of classroom work attendance statistics system based on monitoring camera of the present invention, its work attendance statistics system include as follows
Flow:S101:First, using the method for machine learning, learn a upper half of human body model;S102:Using in S101
The image in the monitor video of classroom is identified in model, so as to calculating the number of student in the classroom;S103:By and learn
The number that should arrive in each classroom is compared in the education administration system of school, so as to calculate late number, number of cutting classes and to class
Rate.
Preferably, the S101:Using the method for machine learning, learn a upper half of human body model, step
It is as follows:
(1.1), positive sample set PS and negative sample set NS is gathered;The positive sample refers to comprising upper half of human body
Image, negative sample refer to the image not comprising upper half of human body;
(1.2), feature is extracted to positive sample and negative sample;
(1.3), after to all positive samples and negative sample extraction feature, the value of feature is write a text file, so
Learnt afterwards using support vector machine method famous in machine learning, upper half of human body can be identified so as to obtain one
Model M.
Preferably, the S102:The image in the monitor video of classroom is identified using the model in S101, from
And calculate the number of student in the classroom;As:The image in the monitor video of classroom is identified using model M, and it is excellent
Change model M, obtain new model M ', so as to calculate the number of student in the classroom;It is as follows:
(2.1), to each two field picture in monitor video, using the model of step S101 middle schools inveterate habit, image is recycled
The famous sliding window technique in identification field, you can identify the specific location of people in the images, detailed process is as follows:
The image of (2.1.1), the size for setting sliding window and positive negative sample are in the same size, i.e. 32 × 32 pixels;For
Each two field picture sliding window in monitor video from left to right, is then slided, often slided on this image from top to bottom
The subgraph consistent with window size is once intercepted, to the subgraph, is judged using the model M of step S101 middle schools inveterate habit
Whether comprising human body in the subgraph, if comprising human body, the value judged is+1, is otherwise -1;Sliding window slides every time
Distance be set as 8 pixels;
(2.1.2), it is set as 8 pixels due to the distance that sliding window slides every time, and in monitor video, on human body
Half body is typically sized to 32 × 32, this can cause to have multiple corresponds to same people by the sliding window that model M judgment value is+1
Body;At this point, the method that maximum inhibits is recycled to be merged to this multiple window so that same human body only corresponds to a cunning
Dynamic window;So as to identify the specific location of human body used in the picture frame;
(2.2), Optimized model M obtains new model M ';Due to that can only identify figure using the method in step (2.1)
As in most of human body, it may appear that missing inspection and flase drop, in order to reduce the situation of missing inspection and flase drop occur, it is necessary to model M into
Row optimization, detailed process are as follows:
(2.2.1), a new classroom monitoring image set S is established, the amount of images which includes is N;
(2.2.2), each image in set S is identified using the method described in step (2.1), then
Artificial examination is carried out to recognition result, the upper half of human body region of missing inspection is added in the positive sample set PS in step (1.1),
The region of flase drop adds in the negative sample set NS in step (1.1), right then again using the method in step (1.2) and (1.3)
Model M is relearned, so as to the model M after being optimized '.
Preferably, the S103:By being compared with the number that should arrive in each classroom in school education administration system, thus
Calculate late number, number of cutting classes and the method to class rate:Using the model M after optimization in step (2.2) ', it recycles
Monitoring image is identified in method described in step (2.1), you can obtains the quantity of human body in the image, thus calculates
To class rate, late number and number of cutting classes, it is as follows:
(3.1), count each classroom arrives class rate:T is the monitoring image in T time point in all classrooms during class hours
Statistical system is uploaded to, monitoring image is identified using the method described in step (2.1), can obtain in classroom
Stranger number NUM1, then compared with the number NUM0 that should arrive in education administration system, you can calculate class rate;
(3.2) during class hours 10-20 minutes after T, then the monitoring image at the time point is uploaded to statistical system, obtained
The number of student NUM2, NUM2-NUM1 at the time point are late number, and NUM0-NUM2 is number of cutting classes.
Beneficial effects of the present invention are:
First, the classroom picture under unmanned state is not required, for establishing the background model in classroom.One is only needed to learn in advance
The upper half of human body model of inveterate habit, you can applied to all classroom environments, greatly reduce the difficulty that system is implemented;
2nd, a threshold value need not be set, for determining whether upper half of human body region.So that the system has preferably
Adaptability;
3rd, can be to arriving class rate, late number, number of cutting classes are counted.
Description of the drawings:
For ease of explanation, the present invention is described in detail by following specific implementations and attached drawing.
Fig. 1 is the flow diagram of the present invention.
Specific embodiment:
Understand to make the object, technical solutions and advantages of the present invention clearer, it is specific below by what is shown in attached drawing
Embodiment describes the present invention.However, it should be understood that these descriptions are merely illustrative, and it is not intended to limit the model of the present invention
It encloses.In addition, in the following description, the description to known features and technology is omitted, to avoid unnecessarily obscuring the present invention's
Concept.
As shown in Figure 1, present embodiment uses following technical scheme:Its work attendance statistics system is included as flowed down
Journey:S101:First, using the method for machine learning, learn a upper half of human body model;S102:Utilize the mould in S101
The image in the monitor video of classroom is identified in type, so as to calculating the number of student in the classroom;S103:By and school
The number that should arrive in each classroom is compared in education administration system, so as to calculate late number, number of cutting classes and to class rate.
Further, the method using machine learning learns a upper half of human body model, and its step are as follows:
1.1st, positive sample set PS and negative sample set NS is gathered;Here positive sample refers to the figure for including upper half of human body
Picture, negative sample refer to the image not comprising upper half of human body, can be classroom, desk, road, computer etc..Positive sample and negative
The image size of sample is all 32 × 32 pixel sizes.In order to enable the model succeeded in school is more applicable for classroom environment, to just
The acquisition of negative sample both is from the monitor video image in classroom.The quantity of wherein positive sample is 3000, the quantity of negative sample
For 20000.The acquisition of positive sample is carried out by the way of artificial, i.e., the classroom image of someone is intercepted by hand comprising on human body
The subgraph in half body region, positive sample set PS will include various postures, clothing, illumination, the human body upper half blocked as far as possible
Body image.The acquisition of negative sample set NS is then to carry out stochastical sampling to the classroom image of nobody to obtain, and contains wall, table
Chair, computer, window etc..
1.2nd, feature is extracted to positive sample and negative sample.The gradient orientation histogram that field of image recognition is employed herein is special
Two merotype features of part of seeking peace.Gradient orientation histogram can be good at describing the local edge of image, local two merotypes
It can be good at describing the textural characteristics of image.Two features are combined, i.e., the corresponding vector order of two features are connected
It connects so that this feature can portray the marginal information and texture information of people well.
1.3rd, after to all positive samples and negative sample extraction feature, the value of feature is write a text file, then
Learnt using support vector machine method famous in machine learning, upper half of human body can be identified so as to obtain one
Model M.
Further, the S102:The image in the monitor video of classroom is identified using the model in S101, from
And calculate the number of student in the classroom;As:The image in the monitor video of classroom is identified using model M, and it is excellent
Change model M, obtain new model M ', so as to calculate the number of student in the classroom;It is as follows:
2.1st, to each two field picture in monitor video, using the model of step S101 middle schools inveterate habit, image is recycled to know
The sliding window technique in other field, you can identify the specific location of people in the images, detailed process is as follows:
2.1.1), the size of setting sliding window and the image of positive negative sample are in the same size, are 32 × 32 pixels.For
Each two field picture (size is typically 1024 × 768 pixels or 640 × 480 pixels) in monitor video, sliding window exists
On the image from left to right, then slided from top to bottom, often slide and once intercept the subgraph consistent with window size,
To the subgraph, judged using the model M of step 1 middle school inveterate habit whether comprising human body in the subgraph, if comprising human body,
The value then judged is+1, is otherwise -1.The distance that sliding window slides every time is set as 8 pixels.
2.1.2), since the distance that sliding window slides every time is set as 8 pixels, and in monitor video, on human body
Half body is typically sized to 32 × 32 or so, so may result in it is multiple by model M judgment value be+1 sliding window correspond to
To same human body.At this point, the method that maximum inhibits is recycled to be merged to this multiple window so that same human body is only right
Answer a sliding window.So as to identify the specific location of human body used in the picture frame.
2.2nd, Optimized model M obtains new model M '.Using the method in step 2.1, can identify big in image
Part human body, but it is present with the situation of missing inspection and flase drop.Missing inspection refers to that the human region in image is not recognized by the system out,
Flase drop refers to it is not that the region of human body is judged as human body in image.Occur to reduce the situation of missing inspection and flase drop so that be
The identification of system is more accurate, it is necessary to be optimized to model M, and detailed process is as follows:
2.2.1 a new classroom monitoring image set S), is established, the amount of images which includes for N (get over by the value of N
Big better, for the value of N for 500), these images will to the greatest extent can with gathering image used in positive negative sample in step 1.1 in the system
Can it is different, and the classroom scene included is rich and varied as far as possible.
2.2.2), each image in set S is identified using the method described in step 2.1, then
Artificial examination is carried out to recognition result, the upper half of human body region of missing inspection is added in the positive sample set PS in step 1.1, by mistake
The region of inspection adds in the negative sample set NS in step 1.1, then again using the method in step 1.2 and 1.3, to model M into
Row relearns, so as to the model M after optimize ', which can greatly reduce missing inspection and the situation appearance of flase drop, make
The identification for the system of obtaining is more accurate.
Further, the S103:By being compared with the number that should arrive in each classroom in school education administration system,
So as to calculate late number, number of cutting classes and the method to class rate:Using the model M after optimization in step (2.2) ', then
Monitoring image is identified using the method described in step (2.1), you can obtain the quantity of human body in the image, thus
Class rate, late number and number of cutting classes are calculated, is as follows:
(3.1), count each classroom arrives class rate:T is the monitoring image in T time point in all classrooms during class hours
Statistical system is uploaded to, monitoring image is identified using the method described in step (2.1), can obtain in classroom
Stranger number NUM1, then compared with the number NUM0 that should arrive in education administration system, you can calculate class rate;
(3.2), during class hours 15 minutes or so after T, then the monitoring image at the time point is uploaded to statistical system, obtained
Number of student NUM2, NUM2-NUM1 to the time point are late number, and NUM0-NUM2 is number of cutting classes.
The basic principles, main features and the advantages of the invention have been shown and described above.The technology of the industry
Personnel are it should be appreciated that the present invention is not limited to the above embodiments, and the above embodiments and description only describe this
The principle of invention, without departing from the spirit and scope of the present invention, various changes and modifications of the present invention are possible, these changes
Change and improvement all fall within the protetion scope of the claimed invention.The claimed scope of the invention by appended claims and its
Equivalent thereof.
Claims (4)
1. a kind of classroom work attendance statistics system based on monitoring camera, it is characterised in that:Its work attendance statistics system includes such as
Lower flow:S101:First, using the method for machine learning, learn a upper half of human body model;S102:Using in S101
Model the image in the monitor video of classroom is identified, so as to calculating the number of student in the classroom;S103:By with
The number that should arrive in each classroom is compared in school's education administration system, so as to calculate late number, number of cutting classes and to class
Rate.
2. a kind of classroom work attendance statistics system based on monitoring camera according to claim 1, it is characterised in that:It is described
S101:Using the method for machine learning, learn a upper half of human body model, its step are as follows:
(1.1), positive sample set PS and negative sample set NS is gathered;The positive sample refers to the figure for including upper half of human body
Picture, negative sample refer to the image not comprising upper half of human body;
(1.2), feature is extracted to positive sample and negative sample;
(1.3), after to all positive samples and negative sample extraction feature, the value of feature is write a text file, is then adopted
Learnt with support vector machine method famous in machine learning, so as to obtain a mould that can identify upper half of human body
Type M.
3. a kind of classroom work attendance statistics system based on monitoring camera according to claim 1, it is characterised in that:It is described
S102:The image in the monitor video of classroom is identified using the model in S101, so as to calculate the student in the classroom
Number;As:The image in the monitor video of classroom is identified using model M, and Optimized model M, obtain new model
M ', so as to calculate the number of student in the classroom;It is as follows:
(2.1), to each two field picture in monitor video, using the model of step S101 middle schools inveterate habit, image identification is recycled
The famous sliding window technique in field, you can identify the specific location of people in the images, detailed process is as follows:
The image of (2.1.1), the size for setting sliding window and positive negative sample are in the same size, i.e. 32 × 32 pixels;For monitoring
Each two field picture sliding window in video from left to right, is then slided, often slided once on this image from top to bottom
The subgraph consistent with window size is intercepted, to the subgraph, judges the son using the model M of step S101 middle schools inveterate habit
Whether comprising human body in image, if comprising human body, the value judged is+1, is otherwise -1;Sliding window slide every time away from
From being set as 8 pixels;
(2.1.2), it is set as 8 pixels due to the distance that sliding window slides every time, and in monitor video, upper half of human body
Be typically sized to 32 × 32, this can cause to have multiple corresponds to same human body by the sliding window that model M judgment value is+1;This
When, the method that maximum inhibits is recycled to be merged to this multiple window so that same human body only corresponds to a sliding window
Mouthful;So as to identify the specific location of human body used in the picture frame;
(2.2), Optimized model M obtains new model M ';Due to that can only be identified using the method in step (2.1) in image
Most of human body, it may appear that missing inspection and flase drop, in order to reduce the situation of missing inspection and flase drop occur, it is necessary to model M carry out it is excellent
Change, detailed process is as follows:
(2.2.1), a new classroom monitoring image set S is established, the amount of images which includes is N;
(2.2.2), each image in set S is identified using the method described in step (2.1), then to knowing
Other result carries out artificial examination, the upper half of human body region of missing inspection is added in the positive sample set PS in step (1.1), flase drop
Region add in negative sample set NS in step (1.1), then again using the method in step (1.2) and (1.3), to model
M is relearned, so as to the model M after being optimized '.
4. a kind of classroom work attendance statistics system based on monitoring camera according to claim 1, it is characterised in that:It is described
S103:By being compared with the number that should arrive in each classroom in school education administration system, so as to calculate late number, cut classes people
Number and the method to class rate:Using the model M after optimization in step (2.2) ', the method described in recycle step (2.1)
Monitoring image to be identified, you can obtain the quantity of human body in the image, thus calculate to class rate, late number and
Cut classes number, be as follows:
(3.1), count each classroom arrives class rate:T uploads the monitoring image in T time point in all classrooms during class hours
To statistical system, monitoring image is identified using the method described in step (2.1), can obtain the student people in classroom
Number NUM1, then compared with the number NUM0 that should arrive in education administration system, you can calculate class rate;
(3.2), during class hours 10-20 minutes after T, then the monitoring image at the time point is uploaded to statistical system, is somebody's turn to do
The number of student NUM2, NUM2-NUM1 at time point are late number, and NUM0-NUM2 is number of cutting classes.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711483705.3A CN108109220A (en) | 2017-12-29 | 2017-12-29 | A kind of classroom work attendance statistics system based on monitoring camera |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711483705.3A CN108109220A (en) | 2017-12-29 | 2017-12-29 | A kind of classroom work attendance statistics system based on monitoring camera |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108109220A true CN108109220A (en) | 2018-06-01 |
Family
ID=62214894
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711483705.3A Pending CN108109220A (en) | 2017-12-29 | 2017-12-29 | A kind of classroom work attendance statistics system based on monitoring camera |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108109220A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109034036A (en) * | 2018-07-19 | 2018-12-18 | 青岛伴星智能科技有限公司 | A kind of video analysis method, Method of Teaching Quality Evaluation and system, computer readable storage medium |
CN110807440A (en) * | 2019-11-15 | 2020-02-18 | 深圳算子科技有限公司 | Method and system for noninductive class face input |
CN112686128A (en) * | 2020-12-28 | 2021-04-20 | 南京览众智能科技有限公司 | Classroom desk detection method based on machine learning |
CN113487758A (en) * | 2021-07-13 | 2021-10-08 | 广东时汇信息科技有限公司 | Classroom arrival rate analysis method |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102354365A (en) * | 2011-09-23 | 2012-02-15 | 上海合合信息科技发展有限公司 | Recognition method and system for accurately recognizing images |
CN104899556A (en) * | 2015-05-14 | 2015-09-09 | 中国农业大学 | System for counting persons in classroom based on image recognition |
CN104985349A (en) * | 2015-07-29 | 2015-10-21 | 高兴 | U-rib weld joint detecting device and U-rib welding system and method |
CN105631410A (en) * | 2015-12-18 | 2016-06-01 | 华南理工大学 | Classroom detection method based on intelligent video processing technology |
CN106228629A (en) * | 2016-07-15 | 2016-12-14 | 合肥师范学院 | Register system in a kind of classroom based on recognition of face |
CN206147769U (en) * | 2016-09-28 | 2017-05-03 | 郑州北斗七星通讯科技有限公司 | Classroom management equipment of registering based on two -dimensional code and face identification |
CN106651844A (en) * | 2016-12-16 | 2017-05-10 | 山东锋士信息技术有限公司 | Apple growing period recognition method based on image analysis |
-
2017
- 2017-12-29 CN CN201711483705.3A patent/CN108109220A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102354365A (en) * | 2011-09-23 | 2012-02-15 | 上海合合信息科技发展有限公司 | Recognition method and system for accurately recognizing images |
CN104899556A (en) * | 2015-05-14 | 2015-09-09 | 中国农业大学 | System for counting persons in classroom based on image recognition |
CN104985349A (en) * | 2015-07-29 | 2015-10-21 | 高兴 | U-rib weld joint detecting device and U-rib welding system and method |
CN105631410A (en) * | 2015-12-18 | 2016-06-01 | 华南理工大学 | Classroom detection method based on intelligent video processing technology |
CN106228629A (en) * | 2016-07-15 | 2016-12-14 | 合肥师范学院 | Register system in a kind of classroom based on recognition of face |
CN206147769U (en) * | 2016-09-28 | 2017-05-03 | 郑州北斗七星通讯科技有限公司 | Classroom management equipment of registering based on two -dimensional code and face identification |
CN106651844A (en) * | 2016-12-16 | 2017-05-10 | 山东锋士信息技术有限公司 | Apple growing period recognition method based on image analysis |
Non-Patent Citations (1)
Title |
---|
曾成斌、刘继乾: "基于图切割和密度聚类的视频行人检测算法", 《模式识别与人工智能》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109034036A (en) * | 2018-07-19 | 2018-12-18 | 青岛伴星智能科技有限公司 | A kind of video analysis method, Method of Teaching Quality Evaluation and system, computer readable storage medium |
CN109034036B (en) * | 2018-07-19 | 2020-09-01 | 青岛伴星智能科技有限公司 | Video analysis method, teaching quality assessment method and system and computer-readable storage medium |
CN110807440A (en) * | 2019-11-15 | 2020-02-18 | 深圳算子科技有限公司 | Method and system for noninductive class face input |
CN110807440B (en) * | 2019-11-15 | 2023-11-10 | 深圳算子科技有限公司 | Classroom face non-sensing input method and system |
CN112686128A (en) * | 2020-12-28 | 2021-04-20 | 南京览众智能科技有限公司 | Classroom desk detection method based on machine learning |
CN113487758A (en) * | 2021-07-13 | 2021-10-08 | 广东时汇信息科技有限公司 | Classroom arrival rate analysis method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109034036B (en) | Video analysis method, teaching quality assessment method and system and computer-readable storage medium | |
CN108109220A (en) | A kind of classroom work attendance statistics system based on monitoring camera | |
CN109359521A (en) | The two-way assessment system of Classroom instruction quality based on deep learning | |
CN108352174B (en) | Electronic device, storage device, and method for image processing | |
CN107408212B (en) | System and method for identifying unwanted photos stored on a device | |
CN106647742B (en) | Movement routine method and device for planning | |
CN109711377B (en) | Method for positioning and counting examinees in single-frame image monitored by standardized examination room | |
CN111507283A (en) | Student behavior identification method and system based on classroom scene | |
CN108027973A (en) | Crowded resolver, crowded analytic method and crowded analysis program | |
CN106204780A (en) | A kind of based on degree of depth study and the human face identification work-attendance checking system and method for cloud service | |
CN109508671A (en) | A kind of video accident detection system and method based on Weakly supervised study | |
CN108647625A (en) | A kind of expression recognition method and device | |
CN108305187A (en) | Dining room based on computer vision is had dinner guidance system | |
CN108229352B (en) | Standing detection method based on deep learning | |
CN110175534A (en) | Teaching assisting system based on multitask concatenated convolutional neural network | |
CN109583373B (en) | Pedestrian re-identification implementation method | |
CN105787522A (en) | Writing attitude evaluation method and writing attitude evaluation system based on handwriting | |
CN110147736A (en) | A kind of repetition anomaly detection method based on posture | |
CN110427815B (en) | Video processing method and device for realizing interception of effective contents of entrance guard | |
CN106709438A (en) | Method for collecting statistics of number of people based on video conference | |
CN105930798A (en) | Tongue image quick detection and segmentation method based on learning and oriented to handset application | |
CN107977645A (en) | A kind of news-video poster map generalization method and device | |
CN110740298A (en) | Distributed classroom discipline behavior detection system, method and medium | |
CN107578424A (en) | A kind of dynamic background difference detecting method, system and device based on space-time classification | |
CN110503092B (en) | Improved SSD monitoring video target detection method based on field adaptation |
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: 20180601 |
|
RJ01 | Rejection of invention patent application after publication |