Method, device and equipment for shooting standing of students in classroom and storage medium
By using a panoramic camera to detect the standing figures within the student area and controlling a close-up camera to capture close-up shots of multiple people or a single person, the problem of unclear shots of multiple people standing in the classroom is solved, improving the shooting effect and the reliability of AI analysis.
Patent Information
- Application Number
- CN202511204861.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-27
- Publication Date
- 2025-12-23
AI Technical Summary
In classroom settings, existing technology cannot clearly capture the entire process of multiple students standing and answering questions, resulting in unclear panoramic images that affect the viewing experience and are not conducive to subsequent AI analysis.
The panoramic camera detects the standing figures in the student area, and when multiple people are detected, the close-up camera is controlled to shoot close-ups of multiple people. The zoom value is calculated using the midpoint of the smallest bounding rectangle to ensure clear shooting. In specific situations, single-person close-up shots are performed or no shots are taken to optimize the shooting effect.
It enables clear filming of students standing and answering questions, improving the viewing quality, meeting the needs of different teaching scenarios, and optimizing the effect of filming multiple people standing.
Smart Images
Figure CN121194059A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of video image processing, and more particularly, to a method and device for shooting students standing in a classroom, an equipment and a storage medium. BACKGROUND
[0002] In a modern education scene, using a panoramic wide-angle fixed-focus camera combined with artificial intelligence (AI) technology to capture students standing up to answer questions in a classroom has become an effective means to improve teaching interactivity and student participation.
[0003] However, in the previous shooting scheme, usually only the teacher or a single student on the podium is shot in close-up, and if multiple people stand in the student area, panoramic shooting is directly used, that is, multiple students answer questions in the student area, or students answer questions at their own positions and the teacher is also in the student area, panoramic shooting is used. In this processing method, when the classroom is relatively large, the panoramic camera cannot clearly shoot the entire process of the students standing up to answer questions. This will result in unclear and poor viewing of the panoramic picture, and these unclear pictures are also not conducive to the use of AI for post-analysis of the classroom. SUMMARY
[0004] To solve the technical problem of clearly shooting the entire process of students standing up to answer questions, the present application provides a method and device for shooting students standing in a classroom, an equipment and a storage medium, and the technical solutions adopted by the present application are as follows.
[0005] In a first aspect, the present application provides a method for shooting students standing in a classroom, comprising:
[0006] receiving a panoramic video picture shot by a panoramic camera; wherein the panoramic camera is used to shoot a panorama of a classroom, and a region in the panoramic video picture is pre-marked as a student area;
[0007] detecting a standing person in the student area in the panoramic video picture and judging the state of the standing person, wherein the state of the standing person includes a standing state and a walking state;
[0008] when multiple people in a standing state are detected, controlling a close-up camera to shoot multiple people in close-up;
[0009] wherein the process of controlling the close-up camera to shoot multiple people in close-up includes:
[0010] finding a minimum circumscribed rectangle of the human body bounding box of all people in a standing state, wherein the minimum circumscribed rectangle has a first width and a first height;
[0011] extracting the midpoint of the minimum circumscribed rectangle;
[0012] Taking the midpoint of the minimum circumscribed rectangle as the midpoint of the single-person close-up shot, taking a single-person close-up shot, obtaining the single-person close-up zoom value at this time, and determining the single-person close-up zoom value as the virtual zoom value;
[0013] Obtaining the proper width and height of the human body inspection frame of the person in the middle of the shot when the zoom value is the virtual zoom value, and recording the second width and the second height;
[0014] If the first width / second width>first height / second height, the actual zoom value=virtual zoom value*first width / second width;
[0015] Otherwise, the actual zoom value=virtual zoom value*first height / second height;
[0016] Controlling the shooting midpoint of the close-up camera to move to the midpoint of the minimum circumscribed rectangle, and shooting based on the actual zoom value.
[0017] In an embodiment, further comprising:
[0018] When it is detected that a person is in a standing state and the teacher is in the podium area, controlling the close-up camera to take a single-person close-up shot of the person in the standing state.
[0019] In an embodiment, further comprising:
[0020] When it is detected that a person is in a standing state and the teacher is not in the podium area, controlling the close-up camera not to take a single-person close-up shot of the student area.
[0021] In an embodiment, the process of the single-person close-up shot comprises:
[0022] Obtaining the midpoint of the human body inspection frame of the person in the standing state;
[0023] Calculating the proportional coefficient of the longitudinal position of the midpoint of the human body inspection frame in the student area, wherein the student area is a pre-defined area;
[0024] Obtaining the maximum zoom value and the minimum zoom value of the close-up camera shooting the student area;
[0025] According to the proportional coefficient, obtaining the single-person close-up zoom value proportionally between the maximum zoom value and the minimum zoom value;
[0026] Controlling the shooting midpoint of the close-up camera to move to the midpoint of the human body inspection frame, and shooting based on the single-person close-up zoom value.
[0027] In an embodiment, the process of the single-person close-up shot further comprises:
[0028] When the single-person close-up shooting is used to shoot only one person subject, a fine-tuning process is further included;
[0029] The fine-tuning process includes:
[0030] According to a preset human body position proportion of the shooting, a human body position proportion point of the human body detection frame of the person in a standing state is obtained;
[0031] A shooting midpoint of the close-up camera is controlled to move to the human body position proportion point;
[0032] A target proportion of the height of the human body detection frame in the height of the picture is obtained;
[0033] A zoom value of the close-up camera is fine-tuned so that the shot picture meets the target proportion.
[0034] In an embodiment, the process of detecting the standing person in the student area in the panoramic video picture and judging the state of the standing person includes:
[0035] A coincidence degree of the detection frame of the standing person in the panoramic video picture detected in the latest N times is calculated;
[0036] When the coincidence degree is in a first preset range, the standing state is judged;
[0037] When the coincidence degree is in a second preset range, the walking state is judged;
[0038] The minimum value of the first preset range is greater than the maximum value of the second preset range.
[0039] In a second aspect, the present application provides a device for shooting standing students in a classroom, comprising:
[0040] A receiving module is configured to receive a panoramic video picture shot by a panoramic camera; wherein the panoramic camera is configured to shoot a panorama of a classroom, and a student area in the panoramic video picture is pre-divided;
[0041] An identifying module is configured to detect a standing person in the student area in the panoramic video picture and judge the state of the standing person, wherein the state of the standing person includes a standing state and a walking state;
[0042] A control module is configured to control a close-up camera to perform a multi-person close-up shooting when it is detected that multiple persons are in a standing state;
[0043] The control module performs the process of controlling the close-up camera to perform the multi-person close-up shooting, including:
[0044] A minimum circumscribed rectangle of a human body inspection frame of all people in a standing state is found, wherein the minimum circumscribed rectangle has a first width and a first height;
[0045] A midpoint of the minimum circumscribed rectangle is extracted;
[0046] A single-person close-up shot is taken using the midpoint of the minimum circumscribed rectangle as a midpoint of the single-person close-up shot, a single-person close-up zoom value at this time is obtained, and the single-person close-up zoom value is determined as a virtual zoom value;
[0047] A proper width and height of a human body inspection frame of a person in the middle of a shot taken when the zoom value is the virtual zoom value are obtained, and are recorded as a second width and a second height;
[0048] If the first width / second width>the first height / second height, an actual zoom value = the virtual zoom value*the first width / second width;
[0049] Otherwise, the actual zoom value = the virtual zoom value*the first height / second height;
[0050] A shooting midpoint of a close-up camera is controlled to move to the midpoint of the minimum circumscribed rectangle, and a shot is taken based on the actual zoom value.
[0051] In an embodiment, the identification module performs the process of detecting a standing person in the student area in the panoramic video shot and judging the state of the standing person, and the process includes:
[0052] A coincidence degree of detection frames of the standing person in the panoramic video shot detected in the latest N times is calculated;
[0053] When the coincidence degree is in a first preset range, it is judged that the person is in a standing state;
[0054] When the coincidence degree is in a second preset range, it is judged that the person is in a walking state;
[0055] The minimum value of the first preset range is greater than the maximum value of the second preset range.
[0056] In a third aspect, the present application provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the method of any of the above embodiments when executing the program.
[0057] In a fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, and the program is executable on a processor to implement the method of any of the above embodiments.
[0058] In the present application, when it is detected that there are multiple students standing in the student area, multiple close-up shooting is enabled, so as to meet the scene that the teacher often lectures on the platform in class, extracts a single student to answer questions, and also supplement the scene that the teacher lets multiple students answer questions together and the scene that the teacher goes down the platform when the students answer questions, and optimize the visual sense of the multiple students standing shooting picture. BRIEF DESCRIPTION OF DRAWINGS
[0059] Figure 1 is a whole flow schematic diagram of the embodiment one of the present application.
[0060] Figure 2 is a whole flow schematic diagram of the embodiment one of the present application.
[0061] Figure 3 is a whole flow schematic diagram of another embodiment one of the present application.
[0062] Figure 4 is a whole structure schematic diagram of the embodiment two of the present application. DETAILED DESCRIPTION
[0063] Exemplary embodiments of the present disclosure will be described hereinafter with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the present disclosure to those skilled in the art.
[0064] It should be noted that the terms "first", "second", and the like in the embodiments of the present application are only to distinguish similar objects, and do not represent a specific order of the objects. Understandably, "first", "second", and the like can be interchanged in a specific order or sequence as appropriate. It should be understood that the objects distinguished by "first", "second", and the like can be interchanged as appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein.
[0065] Embodiment one
[0066] Please refer to Figure 1 , Figure 1 is a flow schematic diagram of a method for shooting students standing in a classroom provided by the embodiment one of the present application, which includes step S110, step S120 and step S130. It should be noted that step S110, step S120 and step S130 are only labels for clearly explaining the correspondence between the embodiments and the drawings, and do not represent the order limitation of each step in the present embodiment. Figure 1
[0067] Step S110, receiving a panoramic video picture captured by a panoramic camera; wherein the panoramic camera is used to capture a panorama of a classroom, and a region in the panoramic video picture is pre-defined as a student region;
[0068] Step S120, detecting a standing person in the student region in the panoramic video picture, and judging a state of the standing person, wherein the state of the standing person includes a standing state and a walking state;
[0069] Step S130, when it is detected that multiple persons are in the standing state, controlling a close-up camera to capture a close-up picture of the multiple persons.
[0070] In a teaching scene, there are generally a panoramic camera specially used to capture a panorama of a classroom including a student region, a plurality of close-up cameras specially used to capture close-up pictures, and a host computer. The host computer is used to analyze pictures captured by the cameras and control the cameras to capture pictures according to analysis results. The camera specially used to capture the panorama of the classroom region generally adopts a wide-angle camera. The method of the embodiment is applied to the host computer of this type.
[0071] In step S110, a panoramic video picture captured by a panoramic camera is received. In the panoramic video picture, a region in the video picture is pre-defined as a student region. Four points of the student region are set, i.e., a left upper corner point P1 (X1, Y1), a right upper corner point P2 (X2, Y2), a right lower corner point (X3, Y3), and a left lower corner point (X4, Y4).
[0072] It should be noted that the XY axis coordinate system used in the embodiment of the application is that the coordinate of the X axis increases from left to right, and the coordinate of the Y axis increases from top to bottom. The XY axis coordinate system used in the embodiment is only an example, and the use of the XY axis coordinate system can be adjusted according to actual needs by those skilled in the art.
[0073] After the panoramic video picture captured by the panoramic camera is received, a standing person in the student region in the panoramic video picture is detected. After the standing person is detected, the coordinate information (x, y, w, h) of the human body bounding box of the standing student is obtained. In the coordinate information, x and y represent the position of the left upper corner point of the bounding box, and w and h represent the width and height of the bounding box, respectively. In general standing detection, as long as a person is standing, the person will be detected. However, the standing person can be actually standing or walking. Therefore, step S120 further judges whether the standing person is in the standing state or the walking state.
[0074] In an embodiment, the process of step S120 includes step S121, step S122, and step S123.
[0075] Step S121: Calculate the overlap of the detection boxes of the latest N detected standing figures in the panoramic video frame;
[0076] Step S122: When the overlap is within a first preset range, it is determined to be a standing state;
[0077] Step S123: When the overlap is within the second preset range, it is determined to be a walking state;
[0078] Among them, the minimum value of the first preset range is greater than the maximum value of the second preset range.
[0079] In this embodiment, the overlap of the detection boxes for the most recent N detected standing figures in the panoramic video image is calculated, for example, 7 times. The results of these 7 detections are compared. If the overlap is within a first preset range (e.g., above 90%), the person is considered standing. If the overlap is within a second preset range (e.g., 20%-90%), the person is considered walking. The remaining overlaps (below 20%) are not included in the results. Based on this, the student area status is determined, such as no one standing, students walking, a single person standing, or multiple people standing.
[0080] In step S130, when multiple people are detected standing, the close-up camera is controlled to perform a close-up shot of the multiple people. This multiple-person shot ensures that each person in the standing position is within the frame of the close-up camera.
[0081] By using close-up shots of multiple people, close-up shots can be taken of all students, thus clearly capturing the entire process of students standing and answering questions.
[0082] Specifically, such as Figure 2 As shown, Figure 2 This is a schematic diagram of the overall process of controlling a close-up camera to shoot close-ups of multiple people. The process of controlling the close-up camera to shoot close-ups of multiple people in step S130 includes: steps S210, S220, S230, S240, S250 and S260.
[0083] Step S210: Find the minimum bounding rectangle of the human body inspection frame for all people in a standing position, wherein the width of the minimum bounding rectangle is a first width and the height is a first height;
[0084] Step S220: Extract the midpoint of the smallest bounding rectangle;
[0085] Step S230: Use the midpoint of the smallest bounding rectangle as the midpoint of the single-person close-up shot, take a single-person close-up shot, obtain the zoom value of the single-person close-up shot at this time, and determine the zoom value of the single-person close-up shot as the virtual zoom value.
[0086] Step S240, the width and height of the human body detection frame of the person in the picture taken when the zoom value is the virtual zoom value are obtained, denoted as the second width and the second height;
[0087] Step S250, if the first width / second width>first height / second height, the actual zoom value=virtual zoom value*first width / second width;
[0088] Otherwise, the actual zoom value=virtual zoom value*first height / second height;
[0089] Step S260, the shooting midpoint of the close-up camera is controlled to move to the midpoint of the minimum circumscribed rectangle, and shooting is performed based on the actual zoom value.
[0090] In step S210, assuming that n persons are standing, the minimum circumscribed rectangle S(xs, ys, ws, hs) of the detection frames of the n persons is obtained, wherein xs and ys represent the position of the top-left corner of the detection frame, and ws and hs represent the width and height of the detection frame, i.e., the first width and the first height of the minimum circumscribed rectangle.
[0091] At this time, the midpoint mids of the minimum circumscribed rectangle is midsx=xs+ws / 2, midsy=ys+hs / 2, and the midpoint position is extracted in step S220.
[0092] For single-person close-up shooting of the student area, there is generally an existing shooting method, and in step S230, assuming that single-person close-up shooting is performed, the existing shooting method is used, the midpoint of the minimum circumscribed rectangle is used as the midpoint of single-person close-up shooting, the single-person close-up zoom value when the existing shooting method is used is obtained, and then the single-person close-up zoom value is determined as the virtual zoom value.
[0093] Assuming that the current virtual zoom value is z0, and assuming that the zoom value of the picture is adjusted to the virtual zoom value z0, the width of the human body detection frame of the person taken, denoted as the second width, is detection frame width w0=(wb-wf)×(midsy-Yf) / (Yb-Yf)+wf; and the height of the human body detection frame of the person taken, denoted as the second height, is detection frame height h0=(hb-hf)×(midsy-Yf) / (Yb-Yf)+hf.
[0094] Wherein, the front row student longitudinal position Yf = (Y1 + Y2) / 2, the back row student longitudinal position Yb = (Y3 + Y4) / 2 in the student area. The front row student shooting zoom value zf, that is, the minimum zoom value for shooting the student area, and the back row student shooting zoom value zb, that is, the maximum zoom value for shooting the student area. The width and height of the human body inspection frame in the picture when shooting the front row student are wf and hf, respectively, and the width and height of the human body inspection frame in the picture when shooting the back row student are wb and hb, respectively.
[0095] The second width and the second height are obtained in step S240.
[0096] Step S250 is a process of obtaining the actual zoom value. In step S250, whether the first width or the first height is used as the basis to determine the actual zoom value is determined by comparison.
[0097] Specifically, if ws / w0>hs / h0, the actual zoom value z = z0*ws / w0; otherwise, the actual zoom value z = z0*hs / h0.
[0098] Step S260 controls the shooting midpoint of the close-up camera to move to the midpoint mids of the minimum circumscribed rectangle, and then shooting is performed using the zoom value z.
[0099] At this time, a close-up picture containing all the standing people in the student area can be shot.
[0100] In the method, when it is detected that multiple people are standing in the student area, multi-person close-up shooting is enabled, so that the scene of the teacher on the platform during class and the scene of extracting a single student to answer a question are met, and the scene of the teacher letting multiple students answer a question and the scene of the teacher going down the platform when the student answers a question are supplemented, and the visual sense of the shooting picture of multiple students standing is optimized.
[0101] In one embodiment, as shown in Figure 3 , Figure 3 is a schematic diagram of the overall flow of another embodiment. The method for shooting the standing students in the classroom further includes step S140.
[0102] Step S140, when it is detected that a person is in a standing state and the teacher is in the platform area, controlling the close-up camera to perform single-person close-up shooting on the person in the standing state.
[0103] This is the case where only one student answers a question. If the teacher is on the platform, it is determined that the standing person is a student, and the close-up camera is controlled to perform single-person close-up shooting.
[0104] In one embodiment, the method for shooting the standing students in the classroom further includes step S150.
[0105] Step S150, when detecting that a person is in a standing state and the teacher is not in the podium area, controlling the close-up camera to not perform a single-person close-up shooting on the student area.
[0106] This is the case where the teacher walks to the student area. When detecting a single-person standing in the student area, if the teacher is under the podium, it is judged that the standing person is most likely the teacher, and no close-up shooting is performed.
[0107] In an embodiment, the process of performing a single-person close-up shooting includes steps S310, S320, S330, S340 and S350.
[0108] Step S310, obtaining the midpoint of the human body detection frame of the person in a standing state;
[0109] Step S320, calculating the proportional coefficient of the midpoint of the human body detection frame in the longitudinal position of the student area, wherein the student area is a pre-defined area;
[0110] Step S330, obtaining the maximum zoom value and the minimum zoom value of the close-up camera shooting the student area;
[0111] Step S340, obtaining a single-person close-up zoom value proportionally between the maximum zoom value and the minimum zoom value according to the proportional coefficient;
[0112] Step S350, controlling the shooting midpoint of the close-up camera to move to the midpoint of the human body detection frame, and performing shooting based on the single-person close-up zoom value.
[0113] This embodiment is a detailed process of performing a single-person close-up shooting. In the single-person close-up shooting, the midpoint (midx0=x0+w0 / 2, midy0=y0+h0 / 2) of the human body detection frame (x0, y0, w0, h0) of the detected single person is taken as the target position of the close-up camera, and the camera is controlled to move to the midpoint, wherein x0 and y0 represent the position of the top-left corner point of the detection frame, and w0 and h0 represent the width and height of the detection frame, respectively.
[0114] Then, the position proportion of the midpoint of the human body detection frame in the longitudinal position of the student area is calculated: (midy0-Yf) / (Yb-Yf). Then, the single-person close-up zoom value z=(zb-zf) x (midy0-Yf) / (Yb-Yf)+zf. At this time, in step S350, the shooting midpoint of the close-up camera is controlled to move to the midpoint of the human body detection frame, and shooting is performed based on the single-person close-up zoom value z.
[0115] Step S230 is to use the above single-person close-up shooting method, and then use the midpoint of the minimum circumscribed rectangle as the midpoint of the single-person close-up shooting to obtain the virtual zoom value.
[0116] In one implementation, when using the single-person close-up shot to shoot a subject with only one person, a fine-tuning process is also included;
[0117] The fine-tuning process includes steps S360, S370, S380, and S390.
[0118] Step S360: Based on the preset human body position ratio, calculate the human body position ratio point of the human body inspection frame for a person in a standing position.
[0119] Step S370: Control the shooting center point of the close-up camera to move to the proportional point of the human body position;
[0120] Step S380: Obtain the target percentage of the height of the human body inspection frame relative to the screen height;
[0121] Step S390: Fine-tune the zoom value of the close-up camera to make the captured image conform to the target proportion.
[0122] This implementation method involves fine-tuning the shooting process after step S350. This fine-tuning process is triggered only when shooting a single person, and is not triggered when determining the virtual zoom value during step S230.
[0123] In step S360, based on the preset human body position ratio, the human body inspection frame of a standing person is calculated to obtain the human body position ratio point. For example, if the shooting position is set at the chest of a single standing person, and the human body frame is (x, y, w, h), then the human body position ratio point is (x+w / 2, y+2*h / 7) (estimated based on the Asian three-quarters height standard). Then, in step S370, the shooting midpoint coordinates of the close-up camera are moved to the human body position ratio point.
[0124] Step S380 obtains the target percentage of the human body inspection frame height to the screen height. For example, if the human body height is set to 1 / 2 of the screen height, the target screen height hp can be calculated. If the current zoom value is z and the screen height is h, then the adjusted zoom value is zp = hp * z / h.
[0125] Example 2
[0126] Corresponding to the method in Example 1, such as Figure 4 As shown, Figure 4 This is a schematic diagram of the overall structure of an embodiment of the present invention. The present invention also provides a device for photographing students standing in the classroom. The device for photographing students standing in the classroom includes: a receiving module 410, an identification module 420, and a control module 430.
[0127] The receiving module 410 is configured to receive a panoramic video picture captured by a panoramic camera; wherein the panoramic camera is configured to capture a panorama of a classroom, and a region in the panoramic video picture is pre-defined as a student region;
[0128] The identifying module 420 is configured to detect a standing person in the student region in the panoramic video picture, and determine a state of the standing person, wherein the state of the standing person includes a standing state and a walking state.
[0129] The control module 430 is configured to control a close-up camera to capture a multi-person close-up when it is detected that multiple persons are in the standing state.
[0130] The control module performs the process of controlling the close-up camera to capture the multi-person close-up, including:
[0131] Obtaining a minimum circumscribed rectangle of the human body bounding boxes of all the persons in the standing state, wherein the minimum circumscribed rectangle has a first width and a first height;
[0132] Extracting a midpoint of the minimum circumscribed rectangle;
[0133] Using the midpoint of the minimum circumscribed rectangle as a midpoint of a single-person close-up, capturing the single-person close-up, obtaining a single-person close-up zoom value at this time, and determining the single-person close-up zoom value as a virtual zoom value;
[0134] Obtaining a proper width and height of a human body bounding box of a person in the middle of a picture captured when the zoom value is the virtual zoom value, and recording the proper width and height as a second width and a second height;
[0135] If the first width / second width> the first height / second height, an actual zoom value = the virtual zoom value* the first width / second width;
[0136] Otherwise, the actual zoom value = the virtual zoom value* the first height / second height;
[0137] Controlling a capturing midpoint of the close-up camera to move to the midpoint of the minimum circumscribed rectangle, and capturing based on the actual zoom value.
[0138] In an embodiment, the identifying module performs the process of detecting the standing person in the student region in the panoramic video picture and determining the state of the standing person, including:
[0139] Calculating a coincidence degree of the latest N times of detecting the human body bounding box of the standing person in the panoramic video picture;
[0140] When the coincidence degree is within a first preset range, determining that the state is the standing state;
[0141] When the coincidence degree is in the second preset range, it is determined that the person is in a walking state;
[0142] The minimum value of the first preset range is greater than the maximum value of the second preset range.
[0143] In an embodiment, the control module is further configured to control the close-up camera to shoot a single-person close-up of the person in the standing state when it is detected that there is only one person in the standing state and the teacher is in the podium area.
[0144] In an embodiment, the control module is further configured to control the close-up camera not to shoot a single-person close-up of the student area when it is detected that there is only one person in the standing state and the teacher is not in the podium area.
[0145] In an embodiment, the process of the single-person close-up shooting comprises:
[0146] obtaining a midpoint of a human body bounding box of the person in the standing state;
[0147] calculating a proportional coefficient of the midpoint of the human body bounding box in the longitudinal position of the student area, wherein the student area is a pre-defined area;
[0148] obtaining a maximum zoom value and a minimum zoom value of the close-up camera shooting the student area;
[0149] deriving a single-person close-up zoom value proportionally between the maximum zoom value and the minimum zoom value according to the proportional coefficient;
[0150] controlling the shooting midpoint of the close-up camera to move to the midpoint of the human body bounding box, and shooting based on the single-person close-up zoom value.
[0151] In an embodiment, the process of the single-person close-up shooting further comprises:
[0152] when the single-person close-up shooting is used to shoot only one person, the process further comprises a fine-tuning process;
[0153] The fine-tuning process comprises:
[0154] obtaining a human body position proportional point of the human body bounding box of the person in the standing state according to a pre-defined human body position proportion of shooting;
[0155] controlling the shooting midpoint of the close-up camera to move to the human body position proportional point;
[0156] obtaining a target proportion of the height of the human body bounding box in the height of the picture;
[0157] fine-tuning the zoom value of the close-up camera so that the shot picture meets the target proportion.
[0158] In the device, when it is detected that multiple students stand in the student area, the multiple-person close-up shooting is enabled, so that the scene of the teacher on the platform and the single student answering the question is met, and the scene of multiple students answering the question and the teacher going down the platform is supplemented, and the visual sense of the multiple-person standing shooting picture is optimized.
[0159] Embodiment three
[0160] The embodiment of the present application also provides a storage medium, which stores computer instructions, and the instructions are executed by a processor to realize the method for shooting the standing students in the classroom according to any of the above embodiments.
[0161] Those skilled in the art can understand that all or part of the steps of the above method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium, and the program is executed to perform the steps of the above method embodiments; and the foregoing storage medium includes a mobile storage device, a random access memory (RAM), a read-only memory (ROM), a magnetic disk or an optical disk and various storage medium capable of storing program codes.
[0162] Alternatively, the integrated unit of the present application can be stored in a computer readable storage medium if it is realized in the form of a software function module and sold or used as an independent product. Based on this understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product, and the computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a terminal or a network device, etc.) to execute all or part of the method embodiments of the present application. The foregoing storage medium includes a mobile storage device, a RAM, a ROM, a magnetic disk or an optical disk and various storage medium capable of storing program codes.
[0163] Corresponding to the above computer storage medium, in one embodiment, a computer device is also provided, which includes a memory, an encoder and a computer program stored in the memory and executable on the encoder, wherein the encoder executes the program to realize the method for shooting the standing students in the classroom and / or the method for shooting the standing students in the classroom according to any of the above embodiments.
[0164] The computer device enables multi-person close-up shooting when detecting that the student area has multiple persons standing, so as to meet the scene of the teacher on the platform teaching and extracting a single student to answer a question, supplement the scene of the teacher letting multiple students answer a question together and the scene of the teacher walking off the platform when the student answers a question, and optimize the visual sense of the multiple student standing shooting picture.
[0165] The technical features of the above-described embodiments can be combined in any manner. To make the description concise, all possible combinations of the technical features in the above-described embodiments are not described, but as long as the combinations of the technical features do not exist in contradiction, they should be considered within the scope of the present disclosure.
[0166] Obviously, the above-described embodiments of the present application are merely examples for clearly illustrating the present application, and are not intended to limit the implementation modes of the present application. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, it is not necessary or possible to exhaust all the implementation modes. Any modification, equivalent replacement and improvement within the spirit and principle of the present application should be included in the protection scope of the claims of the present application.
Claims
1. A method for photographing students standing in a classroom, characterized in that, include: Receive panoramic video footage captured by a panoramic camera; wherein, the panoramic camera is used to capture a panoramic view of the classroom, and a pre-defined area in the panoramic video footage is designated as the student area; Detecting standing figures in the student area of the panoramic video frame and determining the state of the standing figures, wherein the state of the standing figures includes: standing state and walking state; When multiple people are detected standing, control the close-up camera to take close-up shots of the multiple people. The process of controlling the close-up camera to shoot close-up shots of multiple people includes: Find the minimum bounding rectangle of the human body inspection frame for all people in a standing position, where the width of the minimum bounding rectangle is a first width and the height is a first height; Extract the midpoint of the minimum bounding rectangle; Using the midpoint of the smallest bounding rectangle as the midpoint for single-person close-up shooting, single-person close-up shooting is performed, and the zoom value of the single-person close-up at this time is obtained, and the zoom value of the single-person close-up is determined as the virtual zoom value. When the zoom value is a virtual zoom value, obtain the proper width and height of the human body inspection box of the person in the captured image, and record them as the second width and the second height. If the first width / second width > the first height / second height, the actual zoom value = the virtual zoom value * the first width / second width; Otherwise, the actual zoom value = virtual zoom value * first height / second height; The shooting center point of the close-up camera is moved to the center point of the smallest bounding rectangle, and the shot is taken based on the actual zoom value.
2. The method for photographing students standing in a classroom according to claim 1, characterized in that, Also includes: When a person is detected standing and the teacher is in the podium area, the close-up camera is controlled to take a close-up shot of the standing person.
3. The method for photographing students standing in a classroom according to claim 2, characterized in that, Also includes: When a person is detected standing and the teacher is not in the podium area, the close-up camera is controlled not to shoot close-up shots of a single person in the student area.
4. The method for photographing students standing in a classroom according to any one of claims 1-3, characterized in that, The process of shooting a single close-up shot includes: Find the midpoint of the human body inspection frame for a person in a standing position; Calculate the proportional coefficient of the midpoint of the human body examination frame in the longitudinal position of the student area, wherein the student area is a pre-defined area; Obtain the maximum and minimum zoom values of the student area captured by the close-up camera; Based on the scaling factor, the zoom value for a single person close-up is obtained proportionally between the maximum zoom value and the minimum zoom value. The shooting center point of the close-up camera is moved to the center point of the human body inspection frame, and the shot is taken based on the zoom value of the single person close-up.
5. The method for photographing students standing in a classroom according to claim 4, characterized in that, The process of shooting a single close-up shot also includes: When using the aforementioned single-person close-up shooting technique to shoot a subject with only one person, a fine-tuning process is also included. The fine-tuning process includes: Based on the preset proportion of the human body position, the proportion of the human body position in the human body inspection frame for a person in a standing position is calculated. Control the shooting center point of the close-up camera to move to the proportional point of the human body position; Obtain the target percentage of the human body examination frame height relative to the screen height; Fine-tune the zoom value of the close-up camera to ensure that the captured image matches the target proportion.
6. The method for photographing students standing in a classroom according to claim 1, characterized in that, The process of detecting standing figures in the student area of the panoramic video frame and determining the state of the standing figures includes: Calculate the overlap of the detection boxes for the most recent N detected standing figures in the panoramic video frame; When the degree of overlap is within a first preset range, it is determined to be a standing state; When the degree of overlap is within a second preset range, it is determined to be a walking state; Among them, the minimum value of the first preset range is greater than the maximum value of the second preset range.
7. A device for photographing students standing in a classroom, characterized in that, include: The receiving module is used to receive panoramic video images captured by the panoramic camera; wherein, the panoramic camera is used to capture a panoramic view of the classroom, and a region in the panoramic video image is pre-defined as the student area; The recognition module is used to detect standing figures in the student area of the panoramic video image and determine the state of the standing figures, wherein the state of the standing figures includes: standing state and walking state. The control module is used to control the close-up camera to shoot close-up shots of multiple people when it detects that multiple people are standing. The control module executes the process of controlling the close-up camera to shoot close-up shots of multiple people, including: Find the minimum bounding rectangle of the human body inspection frame for all people in a standing position, where the width of the minimum bounding rectangle is a first width and the height is a first height; Extract the midpoint of the minimum bounding rectangle; Using the midpoint of the smallest bounding rectangle as the midpoint for single-person close-up shooting, single-person close-up shooting is performed, and the zoom value of the single-person close-up at this time is obtained, and the zoom value of the single-person close-up is determined as the virtual zoom value. When the zoom value is a virtual zoom value, obtain the proper width and height of the human body inspection box of the person in the captured image, and record them as the second width and the second height. If the first width / second width > the first height / second height, the actual zoom value = the virtual zoom value * the first width / second width; Otherwise, the actual zoom value = virtual zoom value * first height / second height; The shooting center point of the close-up camera is moved to the center point of the smallest bounding rectangle, and the shot is taken based on the actual zoom value.
8. The device for photographing students standing in a classroom according to claim 7, characterized in that, The process by which the recognition module detects standing figures in the student area of the panoramic video frame and determines the state of the standing figures includes: Calculate the overlap of the detection boxes for the most recent N detected standing figures in the panoramic video frame; When the degree of overlap is within a first preset range, it is determined to be a standing state; When the degree of overlap is within a second preset range, it is determined to be a walking state; Among them, the minimum value of the first preset range is greater than the maximum value of the second preset range.
9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-6.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-6.