CN105631410A - Classroom detection method based on intelligent video processing technology - Google Patents
Classroom detection method based on intelligent video processing technology Download PDFInfo
- Publication number
- CN105631410A CN105631410A CN201510970928.7A CN201510970928A CN105631410A CN 105631410 A CN105631410 A CN 105631410A CN 201510970928 A CN201510970928 A CN 201510970928A CN 105631410 A CN105631410 A CN 105631410A
- Authority
- CN
- China
- Prior art keywords
- image
- classroom
- detected
- walks
- detection method
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/20—Education
- G06Q50/205—Education administration or guidance
Abstract
The invention provides a classroom detection method based on an intelligent video processing technology. The method is characterized by comprising a method of counting the number of attendance persons. The attendance person counting method comprises steps of extracting a plurality of material images from a classroom monitoring video, carrying out normalization processing, generating a feature descriptor, adopting an Adaboost algorithm to generate a classifier with the upper body of a student as a recognition target, adopting a scanning sub window to traverse a to-be-detected image, and counting the total attendance persons. The classroom detection method has the advantages that the classroom monitoring cost can be saved; the attendance persons in the classroom can be accurately counted; the counting accuracy is high; automatic teaching quality estimation is realized; the roll call time is saved; and the labor cost is saved.
Description
Technical field
The present invention relates to detection technique field, classroom, more particularly, it relates to a kind of classroom detection method based on intelligent video treatment technology.
Background technology
Intelligent Video Surveillance Technology comes from computer technology, digital image processing techniques and artificial intelligence technology, it utilizes computer vision (ComputerVision, and video analysis (VideoAnalysis CV), VA) method is to a series of analysis of video sequence row, realize the detection of target, location, identification and tracking in dynamic scene, and analyze and judge the behavior of target on this basis, can make a response in time when abnormal conditions occur again thus daily management mission can be completed.
Application based on existing intelligent video analysis is mainly gathered in the abnormality detection of video monitoring, people flow rate statistical etc. Wherein moving object detection, segmentation, recognition and tracking are Railway Project relatively common in the middle of intelligent video analysis research field, are then interesting since the last few years research emphasis problems as behavior understanding and descriptive analysis.
In school, the importance of the classroom situation Dou Shi school controls such as student attendance number, classroom discipline, is the important indicator of TQA. Therefore teacher needs to call the roll to determine attendance on classroom, and school also can arrange supervisor to make an inspection tour classroom; Consume many human resourcess. Present stage, photographic head is generally installed in school classroom and records classroom monitor video, but the effect of classroom monitor video is primarily to and classroom situation is carried out security monitoring; Classroom monitor video is applied to classroom detection field still in blank. Present stage, carry out recognition of face frequently with the Haar+Adaboost strong classifier generated. But the strong classifier that Haar+Adaboost generates is for there being the face at certain angle of inclination to there is check frequency, and Detection results is bad. And the shooting angle of photographic head is fixed in classroom, usual photographic head is all in the oblique upper of student, and the detection object therefore extracted from classroom monitor video is the face at certain angle of inclination. It is found through experiments, adopts the Haar+Adaboost strong classifier generated to be unable to reach the requirement detecting number from the monitor video of classroom.
Summary of the invention
It is an object of the invention to overcome shortcoming of the prior art with not enough, it is provided that a kind of based on intelligent video treatment technology, can save classroom monitoring cost, can accurate statistics goes out number of turning out for work in classroom, statistical accuracy is high, can realize automated teaching quality evaluation, save the roll-call time, save the classroom detection method of human cost.
In order to achieve the above object, the technical scheme is that: a kind of classroom detection method based on intelligent video treatment technology, it is characterised in that include demographic method of turning out for work; Demographic method of turning out for work comprises the steps:
S1 walks, and obtains classroom monitor video, and extract some record images from classroom monitor video from data base; From record image, intercept student region above the waist as material image one, and intercept the region not including student's upper part of the body as material image two; Respectively material image one and material image two are normalized, all equivalently-sized to realize all material image one and material image two;
S2 walks, and respectively each material image one and material image two is divided into some cell factory; Calculate the rectangular histogram of each pixel in each cell factory respectively; It is combined rectangular histogram forming the profiler of material image one and the profiler of material image two respectively;
S3 walks, using the profiler of material image one as positive sample, using the profiler of material image two as negative sample; Adopting Adaboost algorithm to align sample and negative sample learns, generate grader, the identification target of grader is student's upper part of the body;
S4 walks, and obtains the classroom monitor video that classroom to be detected is corresponding, to obtain image to be detected from data base; Image to be detected is carried out pretreatment; Set scanning subwindow, make the initial value that ratio is setting ratio between picture size to be detected and scanning subwindow size; Set people's numerical value as zero;
S5 walks, and adopts scanning subwindow to travel through image to be detected and obtains some scanning subimages; Whether each scanning subimage is the identification target of grader to adopt grader to judge successively: if the identification target of grader, then people's numerical value is from adding one; Otherwise people's numerical value is constant;
S6 walks, it may be judged whether traveled through all setting ratios: if having traveled through all setting ratios, then current people's numerical value is number of always turning out for work, demographics of turning out for work EP (end of program); Otherwise adjusting picture size to be detected and/or scanning subwindow size, making the ratio between picture size to be detected and scanning subwindow size is next setting ratio, and skips to S5 step.
Classroom of the present invention detection method, by school's ubiquity, is used for the classroom monitor video of security monitoring to carry out classroom analysis, is the exploitation to existing resource, can save classroom monitoring cost. Classroom of the present invention detection method can go out number of turning out for work in classroom by accurate statistics, it is achieved automated teaching quality evaluation, saves the roll-call time, saves human cost. Utilize the feature that classroom monitor video camera angle is fixing, adopt the material image learning sample as grader of classroom monitor video extraction, compared with existing grader, the grader that classroom of the present invention detection method trains can realize identifying more effective, quickly and accurately. The identification goal setting of grader is student's upper part of the body, and student refers to the position of more than student's shoulder above the waist; This is owing to classroom middle school student are blocked the latter half by desk, utilizes the part that student is exposed under photographic head in the present invention as much as possible; Compared with only identifying face, identify that student can improve recognition accuracy above the waist.
Further scheme is: classroom detection method also includes classroom discipline detection method; Described classroom discipline detection method comprises the steps:
T1 walks, and obtains time started in classroom and end time, to each two field picture in the classroom monitor video of end time between obtaining from the outset; Gauss hybrid models is adopted to characterize the feature of each pixel in the image corresponding to the time started; Set next frame image as present analysis image;
T2 walks, and is mated with gauss hybrid models by each pixel in present analysis image: if successful match, then judge that this pixel is as background dot; Otherwise judge that this pixel is as foreground point; Adopt present analysis image update gauss hybrid models;
T3 walks, it is judged that whether the time that present analysis image is corresponding is the end time: if so, then skip to T4 step; Otherwise set next frame image as present analysis image, and skip to T2 step;
T4 walks, and deletes background dot respectively to form each frame foreground image in each two field picture; Respectively each frame foreground image binaryzation is obtained each frame black and white foreground image; Adopt function cvFindContours in OpenCV that each frame black and white foreground image is processed, it is thus achieved that the moving target quantity in each frame black and white foreground image;
T5 walks, judge the size between the moving target quantity in each frame black and white foreground image and moving target quantity set limit value respectively: if the moving target quantity >=moving target quantity set limit value in this frame black and white foreground image, then judge that this frame black and white foreground image is as abnormal black and white foreground image; Otherwise judge that this frame black and white foreground image is as normal black and white foreground image;
T6 walks, the frequency of occurrences of the abnormal black and white foreground image of statistics, and adds up continuous some frame black and white foreground images and be the Abnormal lasting of abnormal black and white foreground image, finds out the longest Abnormal lasting; Judge the frequency of occurrences of abnormal black and white foreground image and the longest Abnormal lasting: if the frequency of occurrences >=frequency setting limit value of abnormal black and white foreground image, or the longest Abnormal lasting >=persistent period setting limit value, then judge that this classroom discipline is abnormality; Otherwise judge that this classroom discipline is as normal condition.
Classroom of the present invention detection method also can detect classroom discipline situation, utilizes the image of classroom monitor video to be analyzed judging, testing cost is low, can save classroom monitoring cost, it is not necessary to adopts supervisor to make an inspection tour the monitor mode in classroom, saves human cost. In the detection method of classroom of the present invention, in conjunction with walking about on a large scale etc., specific behavior feature judges, adopt function cvFindContours in OpenCV that each frame black and white foreground image is processed, it is thus achieved that moving target quantity carries out follow-up judgement, can improve judging nicety rate.
Preferred scheme is: in described S2 step, rectangular histogram refers to gradient orientation histogram or edge orientation histogram.
In described S4 step, image to be detected is carried out pretreatment and refers to, undertaken image to be detected reducing noise, compensating photo-irradiation treatment. Due to classroom monitor video record quality uneven, carry out reduce noise, compensate photo-irradiation treatment image to be detected can be optimized, make image to be detected disclosure satisfy that subsequent treatment requirement.
In described S6 step, adjust picture size to be detected and/or scanning subwindow size refers to, adopt one of following three kinds of situations: one, zooming in or out scanning subwindow size, picture size to be detected is constant; Two, reduce picture size to be detected, scan subwindow size constancy; Three, magnified sweep subwindow size, reduces picture size to be detected.
In described T1 step, adopt gauss hybrid models to characterize the feature of each pixel in the image corresponding to the time started and refer to, adopt k Gauss model to characterize the feature of each pixel in the image corresponding to the time started.
In described T4 step, adopt function cvFindContours in OpenCV that each frame black and white foreground image is carried out process to refer to, function cvFindContours in OpenCV is adopted to search moving target profile, delete the area moving target profile less than contour area setting value, calculate residual movement objective contour quantity; Residual movement objective contour quantity is moving target quantity. So process the conversion can avoided because environmental factors causes and cause erroneous judgement.
Compared with prior art, the invention have the advantages that and beneficial effect:
1, classroom of the present invention detection method by school's ubiquity, be used for the classroom monitor video of security monitoring to carry out classroom analysis, be the exploitation to existing resource, classroom monitoring cost can be saved; Classroom of the present invention detection method can go out number of turning out for work in classroom by accurate statistics, it is achieved automated teaching quality evaluation, saves the roll-call time, saves human cost;
2, classroom of the present invention detection method utilizes the feature that classroom monitor video camera angle is fixing, adopt the material image learning sample as grader of classroom monitor video extraction, compared with existing grader, the grader that classroom of the present invention detection method trains can realize identifying more effective, quickly and accurately; The identification goal setting of grader is student's upper part of the body, and student refers to the position of more than student's shoulder above the waist; This is owing to classroom middle school student are blocked the latter half by desk, utilizes the part that student is exposed under photographic head in the present invention as much as possible; Compared with only identifying face, identify that student can improve recognition accuracy above the waist;
3, classroom of the present invention detection method can detect classroom discipline situation, utilizes the image of classroom monitor video to be analyzed judging, testing cost is low, can save classroom monitoring cost, it is not necessary to adopts supervisor to make an inspection tour the monitor mode in classroom, saves human cost.
Accompanying drawing explanation
Fig. 1 is the flow chart of demographic method of turning out for work in the detection method of classroom of the present invention;
Fig. 2 is the flow chart of classroom discipline detection method in the detection method of classroom of the present invention.
Detailed description of the invention
Below in conjunction with accompanying drawing, the present invention is described in further detail with detailed description of the invention.
Embodiment
The present embodiment is based on the classroom detection method of intelligent video treatment technology, including demographic method of turning out for work; Turn out for work the flow process of demographic method as it is shown in figure 1, comprise the steps:
S1 walks, and obtains classroom monitor video, and extract some record images from classroom monitor video from data base; From record image, intercept student region above the waist as material image one, and intercept the region not including student's upper part of the body as material image two; Respectively material image one and material image two are normalized, all equivalently-sized to realize all material image one and material image two;
S2 walks, and respectively each material image one and material image two is divided into some cell factory; Calculate the rectangular histogram of each pixel in each cell factory respectively; It is combined rectangular histogram forming the profiler of material image one and the profiler of material image two respectively;
S3 walks, using the profiler of material image one as positive sample, using the profiler of material image two as negative sample; Adopting Adaboost algorithm to align sample and negative sample learns, generate grader, the identification target of grader is student's upper part of the body;
S4 walks, and obtains the classroom monitor video that classroom to be detected is corresponding, to obtain image to be detected from data base; Image to be detected is carried out pretreatment; Set scanning subwindow, make the initial value that ratio is setting ratio between picture size to be detected and scanning subwindow size; Set people's numerical value as zero;
S5 walks, and adopts scanning subwindow to travel through image to be detected and obtains some scanning subimages; Whether each scanning subimage is the identification target of grader to adopt grader to judge successively: if the identification target of grader, then people's numerical value is from adding one; Otherwise people's numerical value is constant;
S6 walks, it may be judged whether traveled through all setting ratios: if having traveled through all setting ratios, then current people's numerical value is number of always turning out for work, demographics of turning out for work EP (end of program); Otherwise adjusting picture size to be detected and/or scanning subwindow size, making the ratio between picture size to be detected and scanning subwindow size is next setting ratio, and skips to S5 step.
Classroom of the present invention detection method, by school's ubiquity, is used for the classroom monitor video of security monitoring to carry out classroom analysis, is the exploitation to existing resource, can save classroom monitoring cost. Classroom of the present invention detection method can go out number of turning out for work in classroom by accurate statistics, adds up class attendance rate, it is achieved automated teaching quality evaluation, saves the roll-call time, saves human cost. Utilize the feature that classroom monitor video camera angle is fixing, adopt the material image learning sample as grader of classroom monitor video extraction, compared with existing grader, the grader that classroom of the present invention detection method trains can realize identifying more effective, quickly and accurately. The identification goal setting of grader is student's upper part of the body, and student refers to the position of more than student's shoulder above the waist; This is owing to classroom middle school student are blocked the latter half by desk, utilizes the part that student is exposed under photographic head in the present invention as much as possible; Compared with only identifying face, identify that student can improve recognition accuracy above the waist.
Wherein, in described S2 step, rectangular histogram refers to gradient orientation histogram or edge orientation histogram. S2 walks essence and is by image feature selection and extraction; Owing to the data volume of image is sizable, so needing the method utilizing feature extraction to realize the compression of data. The purpose of feature extraction is in that initial data is converted, and obtains the feature that can reflect target essence. In follow-up study, it is possible to replace data by feature. In the present invention, adopt HOG feature to realize, this is because the data useful in a large number of relevant fixed angle face can be collected from the video material of HOG feature, be conducive to subsequent classifier to train.
In S3 step, Adaboost algorithm is a kind of iterative algorithm, its core concept is the grader (Weak Classifier) different for the training of same training set, then these weak classifier set is got up, constitutes a higher final grader (strong classifier). Its algorithm itself realizes by changing data distribution, and whether it is correct according to the classification of sample each among each training set, and the accuracy rate of the general classification of last time, determines the weights of each sample. Give sub classification device by the new data set revising weights to be trained, finally the grader obtained will be trained finally to merge, as last Decision Classfication device every time. Use the grader that Adaboost algorithm is formed can get rid of some unnecessary training data features, and be placed on above the training data of key.
In described S4 step, image to be detected is carried out pretreatment and refers to, undertaken image to be detected reducing noise, compensating photo-irradiation treatment. Due to classroom monitor video record quality uneven, carry out reduce noise, compensate photo-irradiation treatment image to be detected can be optimized, make image to be detected disclosure satisfy that subsequent treatment requirement.
In described S6 step, adjust picture size to be detected and/or scanning subwindow size refers to, adopt one of following three kinds of situations: one, zooming in or out scanning subwindow size, picture size to be detected is constant; Two, reduce picture size to be detected, scan subwindow size constancy; Three, magnified sweep subwindow size, reduces picture size to be detected.
In actual applications, it is also possible to adopt the image except classroom as negative sample.
Classroom of the present invention detection method also includes classroom discipline detection method; The flow process of described classroom discipline detection method is as in figure 2 it is shown, comprise the steps:
T1 walks, and obtains time started in classroom and end time, to each two field picture in the classroom monitor video of end time between obtaining from the outset; Gauss hybrid models is adopted to characterize the feature of each pixel in the image corresponding to the time started; Set next frame image as present analysis image;
T2 walks, and is mated with gauss hybrid models by each pixel in present analysis image: if successful match, then judge that this pixel is as background dot; Otherwise judge that this pixel is as foreground point; Adopt present analysis image update gauss hybrid models;
T3 walks, it is judged that whether the time that present analysis image is corresponding is the end time: if so, then skip to T4 step; Otherwise set next frame image as present analysis image, and skip to T2 step;
T4 walks, and deletes background dot respectively to form each frame foreground image in each two field picture; Respectively each frame foreground image binaryzation is obtained each frame black and white foreground image; Adopt function cvFindContours in OpenCV that each frame black and white foreground image is processed, it is thus achieved that the moving target quantity in each frame black and white foreground image;
T5 walks, judge the size between the moving target quantity in each frame black and white foreground image and moving target quantity set limit value respectively: if the moving target quantity >=moving target quantity set limit value in this frame black and white foreground image, then judge that this frame black and white foreground image is as abnormal black and white foreground image; Otherwise judge that this frame black and white foreground image is as normal black and white foreground image;
T6 walks, the frequency of occurrences of the abnormal black and white foreground image of statistics, and adds up continuous some frame black and white foreground images and be the Abnormal lasting of abnormal black and white foreground image, finds out the longest Abnormal lasting; Judge the frequency of occurrences of abnormal black and white foreground image and the longest Abnormal lasting: if the frequency of occurrences >=frequency setting limit value of abnormal black and white foreground image, or the longest Abnormal lasting >=persistent period setting limit value, then judge that this classroom discipline is abnormality; Otherwise judge that this classroom discipline is as normal condition.
Classroom of the present invention detection method also can detect classroom discipline situation, utilizes the image of classroom monitor video to be analyzed judging, testing cost is low, can save classroom monitoring cost, it is not necessary to adopts supervisor to make an inspection tour the monitor mode in classroom, saves human cost; Dull based on classroom background, fixing feature, utilizes stable background, effectively carries out motion target tracking. In the detection method of classroom of the present invention, in conjunction with walking about on a large scale etc., specific behavior feature judges, adopt function cvFindContours in OpenCV that each frame black and white foreground image is processed, it is thus achieved that moving target quantity carries out follow-up judgement, can improve judging nicety rate.
In described T1 step, adopt gauss hybrid models to characterize the feature of each pixel in the image corresponding to the time started and refer to, adopt k Gauss model to characterize the feature of each pixel in the image corresponding to the time started. Gauss hybrid models mainly has variance and two parameters of average to determine that the study to average and variance is taked different study mechanisms, will be directly influenced the stability of gauss hybrid models, accuracy and convergence. Owing to being that the background extracting to moving target models, it is therefore desirable to variance in gauss hybrid models and two parameter real-time update of average. For improving under busy scene, the Detection results of big and slow moving target, the concept of weights average can be introduced, set up gauss hybrid models real-time update, then pixel is carried out the classification of foreground point and background dot.
In described T4 step, adopt function cvFindContours in OpenCV that each frame black and white foreground image is carried out process to refer to, function cvFindContours in OpenCV is adopted to search moving target profile, delete the area moving target profile less than contour area setting value, calculate residual movement objective contour quantity; Residual movement objective contour quantity is moving target quantity. Search moving target profile and can be to look for facial contour; Judge that whether the area of moving target profile is more than or equal to contour area setting value (such as contour area setting value is 40 pixel * 40 pixels), deletes area less than the moving target profile of contour area setting value afterwards; So process the conversion can avoided because environmental factors causes and cause erroneous judgement.
Above-described embodiment is the present invention preferably embodiment; but embodiments of the present invention are also not restricted to the described embodiments; the change made under other any spirit without departing from the present invention and principle, modification, replacement, combination, simplification; all should be the substitute mode of equivalence, be included within protection scope of the present invention.
Claims (7)
1. the classroom detection method based on intelligent video treatment technology, it is characterised in that include demographic method of turning out for work; Demographic method of turning out for work comprises the steps:
S1 walks, and obtains classroom monitor video, and extract some record images from classroom monitor video from data base; From record image, intercept student region above the waist as material image one, and intercept the region not including student's upper part of the body as material image two; Respectively material image one and material image two are normalized, all equivalently-sized to realize all material image one and material image two;
S2 walks, and respectively each material image one and material image two is divided into some cell factory; Calculate the rectangular histogram of each pixel in each cell factory respectively; It is combined rectangular histogram forming the profiler of material image one and the profiler of material image two respectively;
S3 walks, using the profiler of material image one as positive sample, using the profiler of material image two as negative sample; Adopting Adaboost algorithm to align sample and negative sample learns, generate grader, the identification target of grader is student's upper part of the body;
S4 walks, and obtains the classroom monitor video that classroom to be detected is corresponding, to obtain image to be detected from data base; Image to be detected is carried out pretreatment; Set scanning subwindow, make the initial value that ratio is setting ratio between picture size to be detected and scanning subwindow size; Set people's numerical value as zero;
S5 walks, and adopts scanning subwindow to travel through image to be detected and obtains some scanning subimages; Whether each scanning subimage is the identification target of grader to adopt grader to judge successively: if the identification target of grader, then people's numerical value is from adding one; Otherwise people's numerical value is constant;
S6 walks, it may be judged whether traveled through all setting ratios: if having traveled through all setting ratios, then current people's numerical value is number of always turning out for work, demographics of turning out for work EP (end of program); Otherwise adjusting picture size to be detected and/or scanning subwindow size, making the ratio between picture size to be detected and scanning subwindow size is next setting ratio, and skips to S5 step.
2. the classroom detection method based on intelligent video treatment technology according to claim 1, it is characterised in that also include classroom discipline detection method; Described classroom discipline detection method comprises the steps:
T1 walks, and obtains time started in classroom and end time, to each two field picture in the classroom monitor video of end time between obtaining from the outset; Gauss hybrid models is adopted to characterize the feature of each pixel in the image corresponding to the time started; Set next frame image as present analysis image;
T2 walks, and is mated with gauss hybrid models by each pixel in present analysis image: if successful match, then judge that this pixel is as background dot; Otherwise judge that this pixel is as foreground point; Adopt present analysis image update gauss hybrid models;
T3 walks, it is judged that whether the time that present analysis image is corresponding is the end time: if so, then skip to T4 step; Otherwise set next frame image as present analysis image, and skip to T2 step;
T4 walks, and deletes background dot respectively to form each frame foreground image in each two field picture; Respectively each frame foreground image binaryzation is obtained each frame black and white foreground image; Adopt function cvFindContours in OpenCV that each frame black and white foreground image is processed, it is thus achieved that the moving target quantity in each frame black and white foreground image;
T5 walks, judge the size between the moving target quantity in each frame black and white foreground image and moving target quantity set limit value respectively: if the moving target quantity >=moving target quantity set limit value in this frame black and white foreground image, then judge that this frame black and white foreground image is as abnormal black and white foreground image; Otherwise judge that this frame black and white foreground image is as normal black and white foreground image;
T6 walks, the frequency of occurrences of the abnormal black and white foreground image of statistics, and adds up continuous some frame black and white foreground images and be the Abnormal lasting of abnormal black and white foreground image, finds out the longest Abnormal lasting; Judge the frequency of occurrences of abnormal black and white foreground image and the longest Abnormal lasting: if the frequency of occurrences >=frequency setting limit value of abnormal black and white foreground image, or the longest Abnormal lasting >=persistent period setting limit value, then judge that this classroom discipline is abnormality; Otherwise judge that this classroom discipline is as normal condition.
3. the classroom detection method based on intelligent video treatment technology according to claim 1, it is characterised in that in described S2 step, rectangular histogram refers to gradient orientation histogram or edge orientation histogram.
4. the classroom detection method based on intelligent video treatment technology according to claim 1, it is characterised in that in described S4 step, image to be detected is carried out pretreatment and refers to, is undertaken image to be detected reducing noise, compensating photo-irradiation treatment.
5. the classroom detection method based on intelligent video treatment technology according to claim 1, it is characterized in that, in described S6 step, adjust picture size to be detected and/or scanning subwindow size refers to, adopt one of following three kinds of situations: one, zooming in or out scanning subwindow size, picture size to be detected is constant; Two, reduce picture size to be detected, scan subwindow size constancy; Three, magnified sweep subwindow size, reduces picture size to be detected.
6. the classroom detection method based on intelligent video treatment technology according to claim 2, it is characterized in that, in described T1 step, adopt gauss hybrid models to characterize the feature of each pixel in the image corresponding to the time started to refer to, adopt k Gauss model to characterize the feature of each pixel in the image corresponding to the time started.
7. the classroom detection method based on intelligent video treatment technology according to claim 2, it is characterized in that, in described T4 step, adopt function cvFindContours in OpenCV that each frame black and white foreground image is carried out process to refer to, function cvFindContours in OpenCV is adopted to search moving target profile, delete the area moving target profile less than contour area setting value, calculate residual movement objective contour quantity; Residual movement objective contour quantity is moving target quantity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510970928.7A CN105631410B (en) | 2015-12-18 | 2015-12-18 | A kind of classroom detection method based on intelligent video processing technique |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510970928.7A CN105631410B (en) | 2015-12-18 | 2015-12-18 | A kind of classroom detection method based on intelligent video processing technique |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105631410A true CN105631410A (en) | 2016-06-01 |
| CN105631410B CN105631410B (en) | 2019-04-09 |
Family
ID=56046323
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510970928.7A Active CN105631410B (en) | 2015-12-18 | 2015-12-18 | A kind of classroom detection method based on intelligent video processing technique |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105631410B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108109220A (en) * | 2017-12-29 | 2018-06-01 | 贵州理工学院 | A kind of classroom work attendance statistics system based on monitoring camera |
| CN108257056A (en) * | 2018-01-23 | 2018-07-06 | 余绍志 | A kind of classroom assisted teaching system for the big data for being applied to teaching industry |
| CN111680569A (en) * | 2020-05-13 | 2020-09-18 | 北京中广上洋科技股份有限公司 | Attendance rate detection method, device, equipment and storage medium based on image analysis |
| CN114821399A (en) * | 2022-04-07 | 2022-07-29 | 厦门大学 | Intelligent classroom-oriented blackboard writing automatic extraction method |
| CN111680569B (en) * | 2020-05-13 | 2024-04-19 | 北京中广上洋科技股份有限公司 | Attendance rate detection method, device, equipment and storage medium based on image analysis |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2009301104A (en) * | 2008-06-10 | 2009-12-24 | Chube Univ | Object detecting apparatus |
| CN103699888A (en) * | 2013-12-29 | 2014-04-02 | 深圳市捷顺科技实业股份有限公司 | Human face detection method and device |
| CN104778453A (en) * | 2015-04-02 | 2015-07-15 | 杭州电子科技大学 | Night pedestrian detection method based on statistical features of infrared pedestrian brightness |
-
2015
- 2015-12-18 CN CN201510970928.7A patent/CN105631410B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2009301104A (en) * | 2008-06-10 | 2009-12-24 | Chube Univ | Object detecting apparatus |
| CN103699888A (en) * | 2013-12-29 | 2014-04-02 | 深圳市捷顺科技实业股份有限公司 | Human face detection method and device |
| CN104778453A (en) * | 2015-04-02 | 2015-07-15 | 杭州电子科技大学 | Night pedestrian detection method based on statistical features of infrared pedestrian brightness |
Non-Patent Citations (1)
| Title |
|---|
| 蔡俊: "基于Adaboost算法和背景减法的电梯减法的电梯轿厢内人数检测研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108109220A (en) * | 2017-12-29 | 2018-06-01 | 贵州理工学院 | A kind of classroom work attendance statistics system based on monitoring camera |
| CN108257056A (en) * | 2018-01-23 | 2018-07-06 | 余绍志 | A kind of classroom assisted teaching system for the big data for being applied to teaching industry |
| CN111680569A (en) * | 2020-05-13 | 2020-09-18 | 北京中广上洋科技股份有限公司 | Attendance rate detection method, device, equipment and storage medium based on image analysis |
| CN111680569B (en) * | 2020-05-13 | 2024-04-19 | 北京中广上洋科技股份有限公司 | Attendance rate detection method, device, equipment and storage medium based on image analysis |
| CN114821399A (en) * | 2022-04-07 | 2022-07-29 | 厦门大学 | Intelligent classroom-oriented blackboard writing automatic extraction method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105631410B (en) | 2019-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108805093B (en) | Escalator passenger tumbling detection method based on deep learning | |
| CN102214291B (en) | Method for quickly and accurately detecting and tracking human face based on video sequence | |
| CN104123543B (en) | A kind of eye movement recognition methods based on recognition of face | |
| CN110837784B (en) | Examination room peeping and cheating detection system based on human head characteristics | |
| CN102708361B (en) | Human face collecting method at a distance | |
| CN103279737B (en) | A kind of behavioral value method of fighting based on space-time interest points | |
| CN105469105A (en) | Cigarette smoke detection method based on video monitoring | |
| CN102609724B (en) | Method for prompting ambient environment information by using two cameras | |
| CN108960047B (en) | Face duplication removing method in video monitoring based on depth secondary tree | |
| CN108256462A (en) | A kind of demographic method in market monitor video | |
| CN103605969A (en) | Method and device for face inputting | |
| CN109298785A (en) | A kind of man-machine joint control system and method for monitoring device | |
| CN106886216A (en) | Robot automatic tracking method and system based on RGBD Face datections | |
| CN110728252B (en) | Face detection method applied to regional personnel motion trail monitoring | |
| CN105513053A (en) | Background modeling method for video analysis | |
| CN106709438A (en) | Method for collecting statistics of number of people based on video conference | |
| CN109359549A (en) | A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP | |
| CN107358163A (en) | Visitor's line trace statistical method, electronic equipment and storage medium based on recognition of face | |
| CN105631410A (en) | Classroom detection method based on intelligent video processing technology | |
| Li et al. | Automatic human spermatozoa detection in microscopic video streams based on OpenCV | |
| CN103049748B (en) | Behavior monitoring method and device | |
| CN110991301A (en) | Face recognition method | |
| CN108009480A (en) | A kind of image human body behavioral value method of feature based identification | |
| CN114581819A (en) | Video behavior identification method and system | |
| CN114663835A (en) | Pedestrian tracking method, system, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |