Object recognition method and device, electronic equipment and storage medium

Abstract

The invention discloses an object recognition method and device, electronic equipment and a storage medium. The object identification method comprises the following steps: in response to a received collaborative identification trigger instruction, acquiring a first original image corresponding to a first target area through a teacher camera; segmenting the first original image based on the object type through a teacher camera to obtain a corresponding first type image and a corresponding second type image; when it is detected that a first type of object stands in the first type of image, object standing identification information matched with the first type of object is sent to a student camera through a teacher camera; and focusing the first type of object matched with the object standing identification information through a student camera based on a predetermined target shooting mode to obtain a corresponding object close-up image. According to the embodiment of the invention, the teacher camera assists the student camera in identifying the standing condition of the first type of object, and the accuracy of identifying whether the first type of object stands is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular, to an object recognition method, apparatus, electronic device, and storage medium. Background Art

[0002] For a good recorded lesson, the key lies in whether the director's screen can truly restore the classroom itself. Among them, the teacher calling on a student to answer a question is a very classic and frequently occurring classroom scenario. The teacher randomly selects a student from the classroom, and the student stands up (at this time, the student camera needs to identify from the screen which position there is a standing student, and focus the close-up of the student on the position where this student is located to highlight the student's standing behavior). After the student finishes answering the question, the student sits down, and the teacher continues to give the lecture.

[0003] In the prior art, the student camera starts shooting from the front of the classroom. In the captured image, the images of the students in the front are large, and the images of the students in the back are small. When a student in the back row stands up, there is a probability that the standing student in the back row cannot be recognized, and thus the close-up of the standing student cannot be focused. At the same time, due to the insufficient lens breadth or pixel accuracy of the student camera, the students in the back row or on the side will be connected together, resulting in too low image quality input to the target detection algorithm, further reducing the accuracy of recognizing the standing students. When using a student camera with a wider angle and higher resolution or using a target detection algorithm with higher computing power, the hardware cost is additionally increased. Summary of the Invention

[0004] The present invention provides an object recognition method, apparatus, electronic device, and storage medium to improve the recognition accuracy of the standing situation of the first type of object without increasing the hardware cost.

[0005] According to an aspect of the present invention, an embodiment of the present invention provides an object recognition method applied to a collaborative recognition system. The collaborative recognition system includes: a teacher camera and a student camera. The object recognition method includes:

[0006] In response to a received collaborative recognition trigger instruction, obtain a first original image corresponding to a first target area through the teacher camera;

[0007] Segment the first original image based on the object type through the teacher camera to obtain a corresponding first type of image and a second type of image;

[0008] When it is detected that there is a standing first type of object in the first type of image, send object standing identification information matching the first type of object to the student camera through the teacher camera;

[0009] The student camera focuses on the first type of object that matches the object standing identification information based on a pre-determined target shooting mode, and obtains a corresponding object close-up image.

[0010] According to another aspect of the present invention, there is provided an object recognition device applied to a collaborative recognition system, the collaborative recognition system including: a teacher camera and a student camera; the object recognition device including:

[0011] An image acquisition module, configured to obtain a first original image corresponding to a first target area through the teacher camera in response to a received collaborative recognition trigger instruction;

[0012] An image segmentation module, configured to segment the first original image based on the object type through the teacher camera to obtain a corresponding first type of image and a second type of image;

[0013] A standing recognition module, configured to send object standing identification information matching the first type of object to the student camera through the teacher camera when it is detected that a first type of object stands in the first type of image;

[0014] An object focusing module, configured to focus on the first type of object that matches the object standing identification information based on a pre-determined target shooting mode through the student camera to obtain a corresponding object close-up image.

[0015] According to another aspect of the present invention, there is provided an electronic device, the electronic device including:

[0016] At least one processor; and

[0017] A memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the object recognition method according to any embodiment of the present invention.

[0019] According to another aspect of the present invention, there is provided a computer-readable storage medium storing computer instructions for causing a processor to implement the object recognition method according to any embodiment of the present invention when executed.

[0020] In the technical solution of the embodiment of the present invention, when a collaborative recognition trigger instruction is received by the teacher camera, a first original image corresponding to a first target area is obtained, and then the first original image is segmented based on the object type to obtain a corresponding first type image and a second type image. When it is detected that a first type object stands in the first type image, the teacher camera sends object standing identification information matching the first type object to the student camera. The student camera focuses on the first type object matching the object standing identification information based on the target shooting mode to obtain a corresponding object close-up image, which solves the problem in the prior art that the teacher camera is only used to identify and track the teacher, and the student camera cannot accurately identify the standing of the students in the back row due to insufficient lens breadth or pixel accuracy. Thus, without additional hardware costs, through the cooperation of the teacher camera and the student camera, the standing situation of the first type object can be accurately identified, thereby improving the recognition accuracy of the standing situation of the first type object. At the same time, by adaptively adjusting the shooting mode of the student camera, the standing first type object can be focused and photographed, ensuring that all standing first type objects can be photographed and the standing first type objects are highlighted.

[0021] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. Brief Description of the Drawings

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0023] Figure 1 is a schematic diagram of the hardware deployment of a recording and broadcasting scenario provided according to an embodiment of the present invention;

[0024] Figure 2 is a flowchart of an object recognition method provided according to an embodiment of the present invention;

[0025] Figure 3 is a schematic diagram of the segmentation of a first original image provided according to an embodiment of the present invention;

[0026] Figure 4 is an example diagram of a seating chart of a first type object provided according to an embodiment of the present invention;

[0027] Figure 5It is an exemplary diagram of the coordinate position of a first type of object provided according to an embodiment of the present invention;

[0028] Figure 6 It is a flowchart of another object recognition method provided according to an embodiment of the present invention;

[0029] Figure 7 It is an exemplary diagram of the deployment of a conventional shooting perspective provided according to an embodiment of the present invention;

[0030] Figure 8 It is an exemplary diagram of the deployment of a collaborative shooting perspective provided according to an embodiment of the present invention;

[0031] Figure 9 It is a flowchart of another object recognition method provided according to an embodiment of the present invention;

[0032] Figure 10 It is an exemplary diagram of the adjustment of an actual shooting area provided according to an embodiment of the present invention;

[0033] Figure 11 It is a flowchart of a method for generating a teaching playback video provided according to an embodiment of the present invention;

[0034] Figure 12 It is an example diagram of a first target area provided according to an embodiment of the present invention;

[0035] Figure 13 It is a schematic structural diagram of an object recognition device provided according to an embodiment of the present invention;

[0036] Figure 14 It is a structural block diagram of an electronic device provided according to an embodiment of the present invention. Detailed implementation manners

[0037] In order to enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0038] It should be noted that the terms "first", "second", etc. in the description, claims and above-mentioned drawings of the present invention are used to distinguish similar objects and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order different from those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0039] Figure 1 is a schematic diagram of the hardware deployment of a recording and broadcasting scenario provided according to an embodiment of the present invention. As Figure 1 shown, two cameras (a teacher camera and a student camera respectively), a touch recording and broadcasting host, an interactive display screen, a sound field transformation unit, a directional pick-up microphone, an integrated active speaker and a smart blackboard need to be deployed in the same classroom. Generally speaking, there are 7 pictures (teacher panoramic / teacher close-up / student panoramic / student close-up / computer picture / blackboard writing picture / director picture) in the same classroom, and these pictures are transmitted to the recording and broadcasting host in different ways and further processed by the recording and broadcasting host.

[0040] Among them, the teacher camera is used to shoot the situation of the second type of object area and the front row first type of object area, that is, it can output the panoramic or close-up picture of the second type of object and shoot the situation of the front row first type of object area at the same time; the student camera is used to shoot the situation of the first type of object area, that is, it can output the panoramic or close-up picture of the first type of object at the same time; the smart blackboard can be input into the recording and broadcasting host through the High Definition Multimedia Interface (HDMI) IN method to form a computer picture; the blackboard writing can be input into the recording and broadcasting host through the HDMI IN or camera acquisition method to form a blackboard writing picture; the director picture: refers to a composite picture selected from one or more of the above-mentioned captured pictures as the picture source.

[0041] For the convenience of understanding the following solution, the parameters appearing in this solution are described. Generally speaking, the first type of object refers to the listener, and the second type of object refers to the lecturer. For example, the first type of object can be a student attending the class or other teachers; the second type of object can be a teacher or student giving the lecture.

[0042] It should be noted that the object recognition method in this solution is implemented through a collaborative recognition system. The collaborative recognition system includes a teacher camera and a student camera. Among them, both the teacher camera and the student camera have the function of image shooting, and also have the functions of data storage and data calculation. It can be understood that both the teacher camera and the student camera can be configured with a processor and a memory for storing and processing the acquired images.

[0043] The teacher camera can be understood as a small computing device with a shooting function, having data processing capabilities and data transmission capabilities. Exemplarily, the teacher camera may include logical arithmetic devices such as a Microcontroller Unit (MCU) and a central processing unit (CPU) that have data processing capabilities and / or program execution capabilities.

[0044] The student camera can also be understood as a small computing device with a shooting function, having data processing capabilities and data transmission capabilities. Exemplarily, the student camera may also include logical arithmetic devices such as an MCU and a CPU that have data processing capabilities and / or program execution capabilities.

[0045] In one embodiment, Figure 2 is a flowchart of an object recognition method provided according to an embodiment of the present invention. This embodiment is applicable to accurately recognizing the situation where an object stands. This method can be executed by an object recognition device, and the object recognition device can be implemented in the form of hardware and / or software. The object recognition device can be configured in an electronic device. Among them, the electronic device includes a teacher camera and / or a student camera. As Figure 2 shown, the method includes:

[0046] S110. In response to the received collaborative recognition trigger instruction, obtain a first original image corresponding to a first target area through the teacher camera.

[0047] Among them, the collaborative recognition trigger instruction can be understood as an instruction for controlling the collaborative recognition system to start recognizing the standing situation of the first type of objects. That is to say, after the teacher camera and the student camera in the collaborative recognition system receive the collaborative recognition trigger instruction, they can jointly cooperate to recognize the standing situation of the first type of objects in the target area. The trigger button of the collaborative recognition trigger instruction can be a physical button (such as a button on the hardware) or a virtual button (such as a trigger control on the touch interface). In actual applications, the collaborative recognition trigger instruction can be sent through the recording device, such as being generated in response to triggering the function button of the recording device or by clicking the trigger control on the visualization interface of the recording device. Of course, in the case where the collaborative recognition trigger instruction is generated by clicking the trigger control on the visualization interface of the recording device, the recording device can be a touch device.

[0048] Generally speaking, the target area refers to a space used for teaching. In the actual operation process, the target area can be an indoor space or an outdoor space, as long as the collaborative recognition system is configured in the target area. For example, the target area can be a classroom.

[0049] The first target area refers to the area that needs to be photographed by the teacher camera pre-configured. Exemplarily, the first target area may include the first row of the first type of object area and the podium area. The first original image refers to the unprocessed original picture corresponding to the first target area collected by the teacher camera.

[0050] In the embodiment, when the teacher camera receives the collaborative recognition trigger instruction sent by the recording device, it starts the shooting function of the teacher camera and shoots the actual behavior of each object in the first target area to obtain the first original image corresponding to the first target area.

[0051] S120. Segment the first original image based on the object type by the teacher camera to obtain the corresponding first type image and second type image.

[0052] Among them, the object type refers to different roles that can exist in a target area (for example, listeners or instructors). Generally speaking, in a classroom scenario, the object type can include a first type of object and a second type of object. In actual operations, there can be multiple ways to determine the object type. For example, an image recognition algorithm can be used to recognize the actual behavior of each object in the first original image, and the object type of each object can be determined based on different actual behavior situations; another example is that an image recognition algorithm can be used to recognize the actual orientation of each object in the first original image, and the object type of each object can be determined based on different actual orientation situations; another example is that the object type can be determined according to the distribution of each object in the first original image.

[0053] Exemplarily, in a teaching scenario, the second type of object is generally in a standing state, and the first type of object is generally in a non-standing state. The object type of the object in the standing state can be determined as the second type of object, and the object type of the object in the non-standing state can be determined as the first type of object through methods such as behavior recognition algorithms. Another example is that the teacher's camera is located behind the target area. In the teaching scenario, the first type of object faces the second type of object and has its back to the teacher's camera. The image captured by the first type of object is a back image; correspondingly, the image captured by the second type of object is a front image, and the image captured by the first type of object should be a back image. It can be determined that the object type corresponding to the back image in the first original image is the first type of object, and the object type corresponding to the front image is the second type of object. Another example is that the first type of object is generally densely distributed in rows and columns in the target area, and the second type of object is located alone in the podium area compared to the first type of object. It can be determined that the object with a dense distribution is the first type of object, and the object type corresponding to the object with a dispersed distribution is the second type of object.

[0054] The first type of image refers to an image containing the first type of object, which is the image corresponding to the distribution area of the first type of object; the second type of image refers to an image containing the second type of object, which is the image corresponding to the distribution area of the second type of object.

[0055] In one embodiment, Figure 3 is a schematic diagram of the segmentation of the first original image provided according to an embodiment of the present invention. As Figure 3 shown, when segmenting the first original image 30 based on the object type, the image corresponding to the area where the first type of object is distributed can be used as the first type of image 31, and the image corresponding to the area where the second type of object is distributed can be used as the second type of image 32.

[0056] In the actual operation process, the teacher camera can determine the first type of object and the second type of object included in the first original image 30 based on the object type, determine the area where the first type of object is distributed and the area where the second type of object is distributed, and segment the first original image according to the area where the first type of object is distributed and the area where the second type of object is distributed, to obtain the corresponding first type of image 31 and second type of image 32. In actual applications, it is also possible to only determine the area where the first type of object is distributed, segment the image corresponding to the area where the first type of object is distributed in the first original image into the first type of image 31, and use the remaining image area in the first original image 30 as the second type of image 32.

[0057] S130. When it is detected that there is a first type of object standing in the first type of image, the teacher camera sends the object standing identification information matching the first type of object to the student camera.

[0058] Among them, the object standing identification information is used to represent the standing situation of the first type of object. In the embodiment, the object standing identification information can be represented by the actual coordinate position of each first type of object in the target area, or can also be directly represented by a flag information. Among them, the target area is composed of a first target area and a second target area. For example, if a standing first type of object is located in the third column of the first row in the target area, the object standing identification information can be 13; or, the object standing identification information of a standing first type of object is 1, while the object standing identification information of other non-standing first type of objects in the target area is 0. When the object standing identification information is represented by the actual position of each first type of object in the target area, the object standing identification information can be determined according to the pre-generated object seat map or the pre-acquired actual coordinate position.

[0059] In one embodiment, Figure 4 is an example diagram of a seat map of the first type of object provided according to an embodiment of the present invention. This example takes a seat map of the first type of object with 5 rows and 6 columns as an example. As Figure 4 shown, when the object standing identification information is determined according to the pre-generated object seat map, the student camera can pre-store the seat map of the first type of object, and the object standing identification information is set in the seat map of the first type of object. When there is no first type of object standing, the object standing identification information of each object in the seat map of the first type of object is 0; when there is a first type of object standing in the first type of image, it can trigger the object standing identification information of the standing first type of object in the seat map of the first type of object to be updated to 1. As Figure 4As shown, the first type of standing objects are the first type of object in the first row and second column and the first type of object in the fourth row and fifth column. That is, when a standing first type of object is detected in the first type of image, the teacher camera can update the object standing identification information corresponding to the standing first type of object and send the object standing identification information to the student camera to determine the position of the standing first type of object.

[0060] In one embodiment, Figure 5 is an example diagram of the coordinate position of a first type of object provided according to an embodiment of the present invention. Among them, the number of rows and columns of the first type of object can be not limited and can be determined according to the distribution of the first type of object in the target area. This example takes the coordinate position of the first type of object with 5 rows and 6 columns as an example.

[0061] When there is a standing first type of object in the first type of image, the teacher camera can determine the coordinate position of the standing first type of object. For example, if a standing first type of object is located in the third row and fourth column within the target area, the object standing identification information can be 34, and this coordinate position can be used as the object standing identification information matched with the first type of object. The same coordinate position of the first type of object can be stored in the student camera. After the teacher camera sends the object standing identification information matched with the first type of object to the student camera, the student camera can determine the position of the corresponding first type of object according to the coordinate position.

[0062] In actual applications, the method for detecting the standing of the first type of object in the first original image is not limited, as long as the standing of the first type of object can be recognized. Exemplarily, the standing situation of the first type of object can be determined by means such as camera vision algorithms, displacement algorithms, key point algorithms, etc. In one embodiment, the number of simultaneously standing first type of objects can be multiple, and the teacher camera can simultaneously determine the object standing identification information matched with multiple first type of objects and send it to the student camera.

[0063] S140. The student camera focuses on the first type of object matched with the object standing identification information based on a pre-determined target shooting mode to obtain a corresponding object close-up image.

[0064] Among them, the target shooting mode refers to the shooting mode adopted by the student camera to focus on the first type of object that matches the object standing identification information. Exemplarily, the target shooting mode may include, but is not limited to, a close-up mode, a panoramic mode, and an optimal shooting threshold mode. Among them, the optimal shooting threshold refers to the largest area that the student camera can clearly shoot. In an embodiment, the corresponding target shooting mode is adaptively adjusted according to the actual coordinate positions of all the first type of objects standing within the preset duration in the target area. The preset duration may be a duration set according to user requirements.

[0065] In one embodiment, in a teaching scenario, when a second type of object raises a question, the first type of object can randomly stand up to answer the question. Since multiple first type of objects may not stand up at the same time, it is necessary to determine the actual coordinate positions of all the first type of objects standing within the preset duration and adjust the corresponding target shooting mode to avoid the situation where other standing first type of objects cannot be focused after the first first type of object stands up. The object close-up image refers to a partial shot of the first type of object, which can highlight the first type of object with a standing situation in the target area. Exemplarily, the first type of object can be focused, and a close-up picture can be intercepted to generate an object close-up image; for another example, the focused first type of object can be slightly enlarged (equivalent to giving a close-up shot) or the surrounding background can be blurred to highlight the standing first type of object.

[0066] In an embodiment, since the number of standing first type of objects can be multiple, the first type of objects that match the object standing identification information may be located at various positions in the target area. For example, when the number of the first type of objects that match the object standing identification information is one, the target shooting mode can adopt the close-up mode, such as correspondingly enlarging the standing first type of object. For another example, when the number of the first type of objects that match the object standing identification information is multiple, and the actual coordinate positions of each first type of object are within the optimal shooting threshold of the student camera, the target shooting mode can be the optimal shooting threshold mode, that is, focusing on the first type of object according to the optimal shooting threshold. For another example, when the number of the first type of objects that match the object standing identification information is multiple, and the actual coordinate positions of each first type of object are not within the optimal shooting threshold of the student camera, the target shooting mode can be the panoramic mode to shoot all the standing first type of objects. The student camera can focus on the first type of object that matches the object standing identification information to obtain the corresponding object close-up image.

[0067] Exemplarily, taking the optimal shooting threshold of the student camera as four rows and the area occupied by the actual first type of object as five rows as an example, when the first type of object matching the object standing identification information is at the fourth column of the third row, the close-up mode can be adopted to focus on the first type of object to obtain the corresponding object close-up image. When the first type of objects matching the object standing identification information are respectively at the fourth column of the third row and the second column of the first row, the area occupied by the two standing first type of objects is three rows, which is within four rows, that is, not exceeding the optimal shooting threshold of the student camera. The optimal shooting threshold mode can be adopted to focus on the first type of object to obtain the corresponding object close-up image. When the first type of objects matching the object standing identification information are respectively at the fourth column of the fifth row and the second column of the first row, the area occupied by the two first type of objects is five rows, exceeding the optimal shooting threshold of the student camera. At this time, the panoramic mode can be adopted to focus on the first type of object to obtain the corresponding object close-up image.

[0068] In the technical solution of this embodiment, when the collaborative recognition trigger instruction is received by the teacher camera, the first original image corresponding to the first target area is obtained, and then the first original image is segmented based on the object type to obtain the corresponding first type of image and second type of image. When it is detected that there is a first type of object standing in the first type of image, the teacher camera sends the object standing identification information matching the first type of object to the student camera. The student camera focuses on the first type of object matching the object standing identification information based on the target shooting mode to obtain the corresponding object close-up image, which solves the problem in the prior art that the teacher camera only recognizes the second type of object, and the lens breadth or pixel accuracy of the student camera is insufficient to accurately recognize the standing of the first type of object in the back row. Without additional hardware costs, the teacher camera assists the student camera to recognize the standing situation of the first type of object and the recognition accuracy of whether the first type of object is standing. At the same time, by adaptively adjusting the shooting mode of the student camera, the standing first type of object can be focused and photographed, ensuring that all standing first type of objects can be photographed and the standing first type of objects are highlighted.

[0069] In one embodiment, Figure 6 is a flowchart of another object recognition method provided according to an embodiment of the present invention. On the basis of the above embodiment, when the student camera recognizes that the first type of object is standing, the process of focusing on the first type of object to obtain the corresponding object close-up image is further described. As Figure 6 shown, the method includes:

[0070] S210. In response to the received collaborative recognition trigger instruction, obtain the first original image corresponding to the first target area through the teacher camera, and obtain the second original image corresponding to the second target area through the student camera.

[0071] Among them, the second target area refers to the area that needs to be photographed by the teacher camera pre-configured. Exemplarily, the second target area may include the first type of object area. Exemplarily, when the target area is a classroom, the first target area photographed by the teacher camera may include the first type of object area in the front row. At this time, the second target area may be the remaining first type of object area after removing the first type of object area in the front row included in the first target area. The second original image refers to the unprocessed original picture corresponding to the second target area collected by the student camera.

[0072] In the embodiment, when the teacher camera and the student camera receive the collaborative recognition trigger instruction sent by the recording and broadcasting device, start the shooting functions of the teacher camera and the student camera, and respectively shoot the actual behavior of each object in the first target area and the second target area to obtain the first original image corresponding to the first target area and the second original image corresponding to the second target area.

[0073] S220. Segment the first original image based on the object type through the teacher camera to obtain the corresponding first type of image and second type of image.

[0074] S230. When it is detected that a first type of object stands in the first type of image, send the object standing identification information matching the first type of object to the student camera through the teacher camera.

[0075] S240. When it is detected that a first type of object stands in the second original image, generate the corresponding object standing identification information through the student camera.

[0076] In the embodiment, the first type of object seat map as shown in Figure 4 or the first type of object coordinate position as shown in Figure 5 may also be pre-stored in the student camera. When the student camera generates the corresponding object standing identification information according to the first type of object seat map, when a first type of object stands in the second type of image, the object standing identification information of the standing first type of object in the first type of object seat map can be triggered to be updated to 1. When the student camera generates the corresponding object standing identification information according to the first type of object coordinate position, the coordinate position of the standing first type of object can be determined, and this coordinate position can be used as the object standing identification information matching the first type of object. For example, if a standing first type of object is located in the fourth row and the second column in the target area, the object standing identification information may be 42.

[0077] In actual applications, there is no limitation on the method of detecting the standing of the first type of object in the second original image, as long as the standing of the first type of object can be recognized. The same algorithm as that for detecting the standing of the first type of object in the first original image can be applied. Exemplarily, the method for determining the standing of the first type of object may include, but is not limited to, using camera vision algorithms, displacement algorithms, key point algorithms, etc. In one embodiment, the number of the first type of objects standing simultaneously may be multiple, and the student camera may also determine the object standing identification information corresponding to multiple first type of objects simultaneously.

[0078] S250. The student camera focuses on the first type of objects that match the object standing identification information sent by the teacher camera and the first type of objects that match the object standing identification information generated by the student camera based on a pre-determined target shooting mode, and obtains corresponding object close-up images.

[0079] In the embodiment, the student camera may summarize the object standing identification information sent by the teacher camera and the object standing identification information generated by the student camera, determine the first type of objects that match the object standing identification information sent by the teacher camera and the first type of objects that match the object standing identification information generated by the student camera, and switch the target shooting mode according to the first type of objects that match the object standing identification information. Exemplarily, the corresponding target shooting mode may be adaptively adjusted according to the actual coordinate positions of all the first type of objects standing within a preset time period in the target area.

[0080] In one embodiment, the number of first-type objects that match the object standing identification information sent by the teacher camera and the number of first-type objects that match the object standing identification information generated by the student camera can be multiple, and the first-type objects that match the object standing identification information may be located at various positions in the target area. For example, when the number of first-type objects that match the object standing identification information is one, the target shooting mode can be switched to the close-up mode. Or, when the number of first-type objects that match the object standing identification information is multiple, and the actual coordinate positions of each first-type object are within the optimal shooting threshold of the student camera, correspondingly, the target shooting mode can be the optimal shooting threshold mode. Exemplarily, when the optimal shooting threshold is four rows and the area occupied by the first-type objects is within four rows, the target shooting mode can be adjusted to the optimal shooting threshold mode. Or, when the number of first-type objects that match the object standing identification information is multiple, and the actual coordinate positions of each first-type object are not within the optimal shooting threshold of the student camera, correspondingly, the target shooting mode can be the panoramic mode to capture all the standing first-type objects. Exemplarily, when the optimal shooting threshold is four rows and the area occupied by the first-type objects is outside four rows, the target shooting mode can be the panoramic mode. The student camera can focus on the first-type objects that match the object standing identification information according to the adjusted target shooting mode to obtain corresponding object close-up images.

[0081] It should be noted that the execution order of S230 and S240 is not limited, as long as it is executed before S250.

[0082] The technical solution of this embodiment is that, by responding to the received collaborative recognition trigger instruction, the student camera obtains the second original image corresponding to the second target area. When it is detected that there are first-type objects standing in the second original image, the student camera can generate corresponding object standing identification information. The student camera can focus on the first-type objects that match the object standing identification information sent by the teacher camera and the first-type objects that match the object standing identification information generated by the student camera based on the pre-determined target shooting mode to obtain corresponding object close-up images, realizing the summary and unified focusing of the object standing identification information by the student camera, and preventing the student camera from affecting the live production strategy when transmitting the object standing identification information to the recording and broadcasting device for processing.

[0083] In one embodiment, the object recognition method further includes:

[0084] The student camera sends the object standing identification information to the recording and broadcasting device so that the recording and broadcasting device generates a corresponding teaching playback video based on the object standing identification information and the teaching process.

[0085] Among them, the recording and broadcasting device refers to a device used to generate a corresponding teaching playback video of the teaching process. The recording and broadcasting device may include a touch recording and broadcasting device. The teaching playback video may be a playback video of the timeline trajectory generated according to the actions occurring at each moment in the teaching scenario. The teaching playback video may include a panoramic view of the second type of object, a close-up view of the second type of object, a panoramic view of the first type of object, a close-up view of the first type of object, and a view of the blackboard writing.

[0086] In an embodiment, the student camera may send the object standing identification information to the recording and broadcasting device. The recording and broadcasting device may integrate the panoramic view of the first type of object, the close-up view of the first type of object, the view of the second type of object, and the view of the blackboard writing according to the time sequence of the teaching process and the object standing identification information to generate a teaching playback video. In the teaching scenario, when the second type of object moves and lectures on the podium, the view may be switched to the view of the second type of object to track the view of the second type of object lecturing; when the second type of object writes on the blackboard, the view in the teaching playback video may be correspondingly switched to the view of the blackboard writing; when the second type of object asks a question and the first type of object stands up to answer, the view may be switched to the view of the second type of object according to the object standing identification information to achieve a complete synthesis of the teaching playback video in chronological order and truly restore the classroom itself.

[0087] In one embodiment, before responding to the received collaborative recognition trigger instruction, it further includes:

[0088] Obtain the regional distribution situation corresponding to the target area and the optimal shooting threshold of the student camera; among them, the target area is composed of a first target area and a second target area;

[0089] Deploy the collaborative shooting perspectives of the teacher camera and the student camera based on the regional distribution situation and the optimal shooting threshold.

[0090] Among them, the target area refers to the area that requires the collaborative shooting of the teacher camera and the student camera. In the teaching scenario, the target area may be a classroom. The target area is composed of a first target area and a second target area, that is to say, the shooting perspectives of the teacher camera and the student camera need to cover the target area. The optimal shooting threshold refers to the largest area that the student camera can clearly shoot. Generally speaking, the optimal shooting threshold can be determined by the hardware performance of the student camera itself. It can be understood that the hardware performance of the student camera is proportional to the optimal shooting threshold.

[0091] In an embodiment, the area distribution corresponding to the target area and the optimal shooting threshold value of the student camera can be determined, the student camera perspective is deployed to shoot the second target area according to the optimal shooting threshold value, and the remaining area within the target area (i.e., the first target area) is deployed to shoot the teacher camera perspective.

[0092] In one embodiment, Figure 7 is an example diagram of a conventional shooting perspective deployment provided according to an embodiment of the present invention, such as Figure 7 As shown, the first type of object distribution is taken as five rows as an example, which is a cross-sectional view of the target area. In the conventional shooting angle deployment, the solid line is the shooting angle of the teacher's camera, including the podium area, and the dotted line is the shooting angle of the student camera, including the first type of object area. Figure 8 is an example diagram of a collaborative shooting perspective deployment provided according to an embodiment of the present invention, such as Figure 8 As shown, it is a cross-sectional view of the target area, taking the first type of object distribution as five rows and the optimal shooting threshold value as four rows as an example, the collaborative shooting angles of the teacher camera and the student camera are deployed. Among them, the target area includes the first target area and the second target area.

[0093] The distribution of the area corresponding to the target area can be determined, and the second target area corresponding to the target area can be determined according to the optimal shooting threshold value of the student camera, that is, the second target area is the four rows of first-type object areas close to the teacher camera, and the shooting angle of the student camera is deployed to shoot the second target area. The area of the target area excluding the second target area is used as the first target area, and the shooting angle of the teacher camera is deployed to shoot the first target area, so as to realize the collaborative shooting angle deployment of the teacher camera and the student camera, so as to avoid the phenomenon that the first-type objects in the back row are connected together due to the insufficient lens width or pixel accuracy of the student camera, and then the image quality input to the target detection algorithm is too low, thereby reducing the accuracy of identifying the first-type objects standing. In one embodiment, the object recognition method also includes:

[0094] Respectively obtaining the current position of the first type of object recognized by the student camera as the first current position, and the current position of the first type of object recognized by the teacher camera as the second current position;

[0095] The target shooting mode of the student camera is determined by comparing a to-be-focused area formed by the first current position and the second current position of the student camera with an optimal shooting threshold value of the student camera.

[0096] Among them, the area to be focused can refer to the area that needs to be focused by the student camera, and the first current position and the second current position should be covered in the area to be focused. The area to be focused can be determined according to the minimum row number and the maximum row number among the first current position and the second current position. First, the minimum row number and the maximum row number among the first current position and the second current position can be determined, and the distance between the minimum row number and the maximum row number is determined as the area to be focused. Exemplarily, when the number of the first type of objects is two, and the current positions of the two first type of objects are the first row and the fourth row respectively, the area to be focused is the area occupied by the first type of objects in four rows.

[0097] In the embodiment, the current position of the first type of object recognized by the student camera can be determined as the first current position, and the current position of the first type of object recognized by the teacher camera can be determined as the second current position. The student camera determines the area to be focused according to the first current position and the second current position, and compares the area to be focused with the optimal shooting threshold value to generate a comparison result, and determines the target shooting mode of the student camera according to the comparison result.

[0098] In the actual operation process, the target shooting mode can include a close-up mode, a panoramic mode, and an optimal shooting threshold value mode. When the area to be focused is less than or equal to the optimal shooting threshold value of the student camera, the number of standing first type of objects can be determined. When the number of standing first type of objects is 1, the target shooting mode of the student camera can be determined as the close-up mode; when the number of standing first type of objects is multiple, the target shooting mode of the student camera can be determined as the optimal shooting threshold value mode. When the area to be focused is greater than the optimal shooting threshold value of the student camera, the target shooting mode of the student camera can be determined as the panoramic mode. By dynamically determining the area to be focused, the shooting mode of the adaptive student camera is realized, ensuring that while highlighting the standing first type of objects, all standing first type of objects are captured.

[0099] In one embodiment, the determination method of the current position of the first type of object includes one of the following: determined according to the object standing identification information of the first type of object; determined according to the coordinate position of the first type of object in the target area composed of the first target area and the second target area.

[0100] During the application process, when determining the current position of a first type of object based on the object standing identification information of the first type of object, the object standing identification information of the first type of object can be extracted, and the first type of object seat map corresponding to the object standing identification information can be determined. The position of the first type of object standing corresponding to the identification in the first type of object seat map is determined as the current position of the first type of object. When determining the current position based on the coordinate position of the first type of object in the target area composed of the first target area and the second target area, the pre-stored coordinate position of the first type of object can be determined, and the coordinate position of the first type of object standing corresponding to the identification is determined as the current position of the first type of object.

[0101] In one embodiment, the object recognition method further includes:

[0102] Identifying a second type of object in the second type of image through a teacher camera;

[0103] Tracking the second type of object in the second type of image to obtain the teaching trajectory of the second type of object.

[0104] In an embodiment, after the teacher camera determines the first type of image and the second type of image, it can identify the second type of object in the second type of image. The classroom camera can track the second type of object in the second type of image according to the target tracking algorithm to determine the teaching trajectory of the second type of object. Exemplarily, in a teaching scenario, when the second type of object moves and lectures on the podium, the teacher camera can track the movement of the second type of object and obtain the actions occurring at each moment of the second type of object to generate a teaching trajectory.

[0105] The picture obtained by the teacher camera is segmented, and algorithm operations are respectively performed on the first type of image and the second type of image. While detecting the standing situation of the first type of object in the first type of image, the second type of object can be tracked, realizing the identification of the standing situation of the first type of object in cooperation with the student camera without affecting the functions of the teacher camera itself.

[0106] In one embodiment, Figure 9 is a flowchart of another object recognition method provided according to an embodiment of the present invention. Based on the above embodiment, the process of focusing on the standing first type of object to obtain the corresponding object close-up image is further described. As Figure 9 shown, the method includes:

[0107] S310. In response to the received collaborative recognition trigger instruction, obtain the first original image corresponding to the first target area through the teacher camera.

[0108] S320. Identify the first type of object and the second type of object in the first original image based on the object type through the teacher camera.

[0109] In an embodiment, the teacher camera may determine the first type of object and the second type of object in the first original image according to an image recognition algorithm.

[0110] Exemplarily, in a teaching scenario, the second type of object is generally in a standing state, and the first type of object is generally in a non-standing state. It is possible to determine the object in the standing state in the first original image as the second type of object, and determine the object in the non-standing state as the first type of object. Or, since the teacher camera is located at the back of the classroom, the image of the second type of object is a front image, and the image of the first type of object is a back image. It is possible to determine the object corresponding to the back image in the first original image as the first type of object, and determine the object corresponding to the front image in the first original image as the second type of object. Or, the first type of object is generally densely distributed in rows and columns in the classroom, and the second type of object is located alone in the podium area compared with the first type of object. It is possible to determine the object with a dense distribution in the first original image as the first type of object, and the object with a scattered distribution as the second type of object.

[0111] S330. Obtain the corresponding first type of area and second type of area respectively based on the distribution area of the first type of object and the distribution area of the second type of object.

[0112] In an embodiment, the distribution area of the first type of object and the distribution area of the second type of object may be determined respectively. The distribution area of the first type of object is used as the first type of area, and the distribution area of the second type of object is used as the second type of area. In the actual operation process, it is also possible to first determine the area where the second type of object is located, divide the area where the second type of object is located into the second type of area, and divide the remaining area in the first original image into the first type of area.

[0113] S340. Segment the first original image according to the first type of area and the second type of area to obtain the corresponding first type of image and second type of image.

[0114] In an embodiment, the first original image may be segmented by an image segmentation technique. After determining the first type of area and the second type of area, segment the first original image according to the first type of area and the second type of area. The area corresponding to the first type of area is used as the first type of image, and the area corresponding to the second type of area is used as the second type of image.

[0115] S350. When it is detected that there is a standing first type of object in the first type of image, send the object standing identification information matching the first type of object to the student camera through the teacher camera.

[0116] S360. Dynamically adjust the actual shooting area of the student camera based on a pre-determined target shooting mode through the student camera.

[0117] Among them, the actual shooting area may refer to the actual shooting range of the student camera. When the actual shooting area is small, the corresponding first-type object image is larger; correspondingly, when the actual shooting area is large, the corresponding first-type object image is smaller.

[0118] In an embodiment, the student camera can change the target shooting mode and adjust the actual shooting area according to the target shooting mode. During the application process, a pan-tilt (i.e., a rotating shaft) can be set in the student camera, and the actual shooting area can be adjusted by rotating the pan-tilt. During the actual operation process, when the target shooting mode is the close-up mode, the student camera can place the first-type object at the center of the screen and adjust the actual shooting area of the student camera. When the target shooting mode is the optimal shooting threshold mode, the current position of the first target object can be determined, the area to be focused can be determined, and the actual shooting area of the student camera can be adjusted according to the area to be focused. When the target shooting mode is the panoramic mode, the screen of the student camera can be zoomed out, and all the first target objects can be placed in the actual shooting area so that all the first target objects can be captured by the student camera.

[0119] In one embodiment, Figure 10 is an example diagram of the adjustment of the actual shooting area provided according to an embodiment of the present invention. As Figure 8 shown, it is the original deployment method of the shooting perspective of the student camera. In the case where there is no first-type object standing, the shooting perspective of the student camera is the rear row area in the first-type object. When there is a standing first-type object in the first row near the podium area, the target shooting mode of the student camera can adopt the optimal shooting threshold mode, and move the shooting perspective of the student camera to Figure 10 the perspective shown to capture all the standing first-type objects. As Figure 10 shown, the current shooting perspective of the student camera is the area of the first four rows of first-type objects near the podium area. At this time, the actual shooting area of the student camera is the area of the first four rows of first-type objects near the podium area.

[0120] S370. Focus on the first-type objects in the actual shooting area that match the object standing identification information to obtain corresponding object close-up images.

[0121] In an embodiment, the position of the first type of object that matches the object standing identification information in the actual shooting area can be determined, and the first type of object can be focused on. During the focusing process, the surrounding background can be appropriately blurred to highlight the first type of object that matches the object standing identification information; alternatively, a close-up picture of the first type of object can be intercepted to obtain a corresponding object close-up image.

[0122] The technical solution of this embodiment, in response to the received collaborative recognition trigger instruction, obtains the first original image corresponding to the first target area through the teacher camera, and then the teacher camera identifies the first type of object and the second type of object in the first original image based on the object type, and respectively obtains the corresponding first type area and the second type area based on the distribution area of the first type of object and the distribution area of the second type of object, so as to achieve the accurate division of the first type area and the second type area. When it is detected that there is a first type of object standing in the first type of image, the teacher camera sends the object standing identification information that matches the first type of object to the student camera, and at the same time the student camera dynamically adjusts the actual shooting area of the student camera based on the pre-determined target shooting mode, and focuses on the first type of object that matches the object standing identification information in the actual shooting area to obtain a corresponding object close-up image, so as to realize the dynamic adjustment of the actual shooting area of the student camera and ensure that all standing first type of objects can be photographed.

[0123] In one embodiment, Figure 11 is a flowchart of a method for generating a teaching playback video according to an embodiment of the present invention. In this embodiment, taking the target area as the classroom, the recording and broadcasting device as the recording and broadcasting host, the first type of object as the student, and the second type of object as the teacher as an example, the collaborative recognition trigger instruction is generated by the recording and broadcasting device, the teacher camera and the student camera are deployed in a collaborative perspective, and the teaching playback video is generated by the recording and broadcasting device. As Figure 11 shown, the process of the teaching playback video in this embodiment includes the following steps:

[0124] S410. Deploy the collaborative shooting perspectives of the teacher camera and the student camera according to the area distribution of the classroom and the optimal shooting threshold value of the student camera.

[0125] Among them, the classroom is divided into a first target area and a second target area. The first target area is the area that the teacher camera can shoot; the second target area is the area that the student camera can shoot.

[0126] S420. Trigger the collaborative mode button of the recording and broadcasting host to generate a collaborative recognition trigger instruction.

[0127] Among them, the collaborative mode button is used to trigger the generation of a collaborative recognition trigger instruction.

[0128] S430. The live recording host sends the collaborative recognition trigger instruction to the teacher camera and the student camera respectively.

[0129] S440. In response to the collaborative recognition trigger instruction, the teacher camera cuts the image of the first target area into a first type of image and a second type of image.

[0130] In one embodiment, Figure 12 is an example diagram of a first target area provided according to an embodiment of the present invention. Among them, the first type of image may include the front row area of the first type of object, and the second type of image includes the podium area (including the blackboard, podium, and second type of object, etc.).

[0131] S450. When the teacher camera detects that there is a first type of object standing in the first type of image, it sends the object standing identification information matching the first type of object to the student camera.

[0132] S460. The student camera waits for the object standing identification information sent by the teacher camera.

[0133] S470. When the student camera detects that there is a first type of object standing in the second type of image, it generates the object standing identification information matching the first type of object.

[0134] S480. The student camera focuses on the matching first type of object according to the object standing identification information to obtain the corresponding object close-up image.

[0135] S490. The student camera sends the object standing identification information to the live recording host so that the live recording host generates a corresponding teaching playback video based on the object standing identification information and the teaching trajectory of the teacher.

[0136] It should be noted that the execution order of S460 and S470 is not limited, as long as it is executed before S250.

[0137] The technical solution of this embodiment, through the deployment of the collaborative shooting perspectives of the teacher camera and the student camera, avoids the phenomenon that due to the insufficient lens breadth or pixel accuracy of the student camera, the first type of objects in the back row are connected together, and then the image quality input to the target detection algorithm is too low, affecting the accuracy of recognizing the standing of the first type of object.

[0138] In one embodiment, Figure 13 is a schematic structural diagram of an object recognition device provided according to an embodiment of the present invention. As Figure 13 shown, the device includes: an image acquisition module 131, an image segmentation module 132, a standing recognition module 133, and an object focusing module 134.

[0139] Among them, the image acquisition module 131 is configured to obtain a first original image corresponding to a first target area through a teacher camera in response to a received collaborative recognition trigger instruction;

[0140] The image segmentation module 132 is configured to segment the first original image based on object types through the teacher camera to obtain corresponding first-type images and second-type images;

[0141] The standing recognition module 133 is configured to send object standing identification information matching the first-type object to the student camera through the teacher camera when it detects that a first-type object stands in the first-type image;

[0142] The object focusing module 134 is configured to focus on the first-type object matching the object standing identification information through the student camera based on a pre-determined target shooting mode to obtain a corresponding object close-up image.

[0143] In one embodiment, the object recognition device further includes:

[0144] The second image acquisition module is configured to obtain a second original image corresponding to a second target area through the student camera in response to a received collaborative recognition trigger instruction;

[0145] The second standing recognition module is configured to generate corresponding object standing identification information through the student camera when it detects that a first-type object stands in the second original image;

[0146] Correspondingly, the object focusing module 134 is further configured to:

[0147] Focus on the first-type object matching the object standing identification information sent by the teacher camera and the first-type object matching the object standing identification information generated by the student camera through the student camera based on a pre-determined target shooting mode to obtain a corresponding object close-up image.

[0148] In one embodiment, the object recognition device further includes:

[0149] The video generation module is configured to send the object standing identification information to a recording and broadcasting device through the student camera, so that the recording and broadcasting device generates a corresponding teaching playback video based on the object standing identification information and the teaching process.

[0150] In one embodiment, the object recognition device further includes:

[0151] The distribution situation acquisition module is configured to obtain the regional distribution situation corresponding to the target area and the optimal shooting threshold value of the student camera; wherein, the target area is composed of a first target area and a second target area;

[0152] A perspective deployment module for deploying the collaborative shooting perspectives of teacher cameras and student cameras based on the regional distribution and the optimal shooting threshold value.

[0153] In one embodiment, the object recognition device further includes:

[0154] A current position acquisition module for respectively acquiring the current position of the first type of object recognized by the student camera as the first current position, and the current position of the first type of object recognized by the teacher camera as the second current position;

[0155] A shooting mode determination module for determining the target shooting mode of the student camera based on the comparison result between the to-be-focused area formed by the first current position and the second current position and the optimal shooting threshold value of the student camera through the student camera.

[0156] In one embodiment, the determination method of the current position of the first type of object in the object recognition device includes one of the following: determined according to the object standing identification information of the first type of object; determined according to the coordinate position of the first type of object in the target area formed by the combination of the first target area and the second target area.

[0157] In one embodiment, the image segmentation module 132 includes:

[0158] An object recognition unit for recognizing the first type of object and the second type of object in the first original image based on the object type through the teacher camera;

[0159] An area determination unit for respectively obtaining the corresponding first type of area and the second type of area based on the distribution area of the first type of object and the distribution area of the second type of object;

[0160] An image segmentation unit for segmenting the first original image according to the first type of area and the second type of area to obtain the corresponding first type of image and the second type of image.

[0161] In one embodiment, the object focusing module 134 includes:

[0162] A shooting area determination unit for dynamically adjusting the actual shooting area of the student camera based on the pre-determined target shooting mode through the student camera;

[0163] An object focusing unit for focusing on the first type of object that matches the object standing identification information in the actual shooting area to obtain the corresponding object close-up image.

[0164] In one embodiment, the object recognition device further includes:

[0165] A second object recognition module, configured to recognize a second type of object in a second type of image through a teacher camera;

[0166] A second object tracking module, configured to track a second type of object in a second type of image to obtain a teaching trajectory of the second type of object.

[0167] The object recognition device provided by an embodiment of the present invention can execute the object recognition method provided by any embodiment of the present invention, and has function modules and beneficial effects corresponding to the execution of the method.

[0168] In one embodiment, Figure 14 is a block diagram of a structure of an electronic device provided according to an embodiment of the present invention, as Figure 14 shown, which shows a schematic structural diagram of an electronic device 10 that can be used to implement the embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, a personal digital processor, a cellular phone, a smart phone, a wearable device (such as a helmet, glasses, a watch, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present invention described and / or claimed herein. Exemplarily, the electronic device 10 can be a teacher camera and / or a student camera.

[0169] As Figure 14 shown, the electronic device 10 includes at least one processor 11, and a memory communicatively connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc. Among them, the memory stores a computer program executable by the at least one processor. The processor 11 can execute various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. The input / output (I / O) interface 15 is also connected to the bus 14.

[0170] Multiple components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disc, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0171] The processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the object recognition method.

[0172] In some embodiments, the object recognition method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the object recognition method described above can be executed. Alternatively, in other embodiments, the processor 11 can be configured to execute the object recognition method by any other suitable means (e.g., by means of firmware).

[0173] The various embodiments of the systems and technologies described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor, and can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0174] A computer program for implementing the method of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, a special purpose computer, or other programmable data processing device, such that the computer programs, when executed by the processor, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The computer programs may be executed entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine, or entirely on the remote machine or server.

[0175] In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium may include, but is not limited to, electronics, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium may be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0176] In order to provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0177] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected with each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), blockchain network, and the Internet.

[0178] A computing system can include a client and a server. The client and the server are generally remote from each other and typically interact via a communication network. The client-server relationship is created by computer programs running on respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system and solves the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services.

[0179] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is made herein.

[0180] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. An object recognition method, characterized in that, Applied to a collaborative recognition system, the collaborative recognition system includes: a teacher camera and a student camera; the object recognition method includes: In response to the received collaborative recognition trigger instruction, acquiring a first original image corresponding to the first target area through the teacher camera; Segmenting the first original image based on the object type by the teacher camera to obtain corresponding first type images and second type images; When it is detected that there is a first type of object standing in the first type of image, sending object standing identification information matching the first type of object to the student camera through the teacher camera; The student camera focuses on the first type of object matching the object standing identification information based on a predetermined target shooting mode to obtain a corresponding object close-up image.

2. The method according to claim 1, wherein The method further comprises: In response to the received collaborative recognition trigger instruction, acquiring a second original image corresponding to the second target area through the student camera; When it is detected that there is a first type of object standing in the second original image, generating corresponding object standing identification information through the student camera; Accordingly, focusing the first type of object matching the object standing identification information by the student camera based on a predetermined target shooting mode to obtain a corresponding close-up image of the object includes: The student camera focuses on the first type of object matching the object standing identification information sent by the teacher camera and the first type of object matching the object standing identification information generated by the student camera based on a predetermined target shooting mode to obtain a corresponding object close-up image.

3. The method according to claim 2, wherein The method further comprises: The object standing identification information is sent to the recording and broadcasting device through the student camera, so that the recording and broadcasting device generates a corresponding teaching playback video based on the object standing identification information and the teaching process.

4. The method according to claim 2, wherein Before responding to the received collaborative identification trigger instruction, the method further includes: Obtaining the area distribution corresponding to the target area and the optimal shooting threshold value of the student camera; wherein the target area is composed of the first target area and the second target area; The collaborative shooting angles of the teacher camera and the student camera are deployed based on the area distribution and the optimal shooting threshold value.

5. The method according to claim 2, characterized in that, The method further comprises: Respectively obtaining the current position of the first type of object recognized by the student camera as the first current position, and the current position of the first type of object recognized by the teacher camera as the second current position; The target shooting mode of the student camera is determined by comparing the target shooting mode of the student camera with the optimal shooting threshold value of the student camera based on the comparison result between the area to be focused formed by the first current position and the second current position.

6. The method according to claim 5, wherein The current position of the first type object is determined in one of the following ways: based on the object standing identification information of the first type object; based on the coordinate position of the first type object in the target area formed by the combination of the first target area and the second target area.

7. The method according to any one of claims 1-6, characterized in that, Segmenting the first original image based on object types by the teacher camera to obtain corresponding first-type images and second-type images includes: Identifying first-type objects and second-type objects in the first original image based on object types by the teacher camera; Respectively obtaining corresponding first-type regions and second-type regions based on the distribution regions of the first-type objects and the distribution regions of the second-type objects; Segmenting the first original image according to the first-type regions and the second-type regions to obtain corresponding first-type images and second-type images.

8. The method according to any one of claims 1-6, characterized in that Focusing on first-type objects matching the object standing identification information based on a pre-determined target shooting mode by the student camera to obtain corresponding object close-up images includes: Dynamically adjusting the actual shooting area of the student camera based on a pre-determined target shooting mode by the student camera; Focusing on first-type objects matching the object standing identification information in the actual shooting area to obtain corresponding object close-up images.

9. The method according to any one of claims 1 to 6, characterized in that, The method further includes: Identifying second-type objects in the second-type images by the teacher camera; Tracking the second-type objects in the second-type images to obtain the teaching trajectories of the second-type objects.

10. An object recognition device, characterized in that, Applied to a collaborative recognition system, the collaborative recognition system includes: a teacher camera and a student camera; the object recognition device includes: An image acquisition module, configured to obtain a first original image corresponding to a first target area through the teacher camera in response to a received collaborative recognition trigger instruction; An image segmentation module, configured to segment the first original image based on object types by the teacher camera to obtain corresponding first-type images and second-type images; A standing recognition module, configured to send object standing identification information matching the first-type objects to the student camera through the teacher camera when it is detected that a first-type object stands in the first-type images; An object focusing module, configured to focus on first-type objects matching the object standing identification information based on a pre-determined target shooting mode by the student camera to obtain corresponding object close-up images.

11. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the object recognition method according to any one of claims 1-9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and the computer instructions are used to implement the object recognition method according to any one of claims 1-9 when executed by a processor.

Images