Intelligent scoring system for making human lung and external gas exchange model experiment

Abstract

The invention discloses an intelligent scoring system for an experiment of making a human lung and external gas exchange model, and the system comprises a hardware module which is composed of a student tablet with a camera and a server with high computing power. The student tablet portion with the camera is composed of a camera unit, an interactive touch input unit and a display unit, and the server portion is composed of an image decoding unit, a deep network unit, a logic analysis unit and an image coding unit. According to the intelligent scoring system for making the human lung and external gas exchange model experiment, YOLO has the characteristics of being fine, small and rapid, the time for service logic codes of the human lung model experiment is very short, the scoring condition of experiment operation can be directly given in real time in the experiment process of students, and the experiment efficiency is improved. And the human lung model experiment is used as a part of the physical and chemical student experiment, so that the integrity of intelligent scoring of the physical and chemical student experiment is also improved.

Description

Technical Field

[0001] The invention relates to the field of teaching technology, and in particular to an intelligent scoring system for making a model experiment of gas exchange between human lungs and the outside world. Background Art

[0002] The three subjects of junior high school physics, chemistry and biology (physics, chemistry, and biology) include a total of 70+ separate experiments. Making a model of gas exchange between human lungs and the outside world is one of the biological experiments. The practical operation of physics, chemistry and biology experiments has been planned to be included in the scope of the junior high school entrance examination, and junior high schools have also begun to invest in the human and material costs of the practical operation courses of physics, chemistry and biology experiments. The practical operation of experiments emphasizes the standardization of operations and focuses on the process. However, a teacher's attention is limited. At most, he can guide (invigilate) the experimental operations of 2 to 3 students at the same time. Especially during the practical operation examination, this undoubtedly consumes a lot of human resources. There are technologies that apply AI to physics, chemistry and biology experiments. Target detection based on YOLO is a fairly mature AI technology. In industrial applications, using mature technologies to solve problems has always been the best way to use products. YOLO simplifies the target detection problem into a regression problem, starting directly from image pixels to obtain the box and classification probability.

[0003] After searching, it is found that the AI ​​of physical, chemical and biological experiments in the prior art is typically such as the publication number CN111314390A, which is an AI classroom system of physical, chemical and biological experiments based on image recognition, including a laboratory table, an intelligent analysis device for the laboratory classroom, and an external device. The laboratory table includes a student terminal and a network camera. The external device includes a teacher terminal, a teacher mobile terminal, and a student terminal. The network camera of the laboratory table communicates with the intelligent analysis device of the laboratory classroom via the network, and collects the video of the student's operation experiment through the network camera. The intelligent analysis device of the laboratory classroom includes a receiving unit, an intelligent analysis unit, and a control unit. The intelligent analysis unit receives the experimental step judgment request sent by the student terminal of the laboratory table and the information of the relevant network camera from the receiving unit, obtains the current experimental table experimental status video data shot by the network camera for image processing and AI analysis, and then determines whether the current step result is correct. Its main feature is that when the artificial intelligence analysis unit cannot determine whether the steps submitted by the current student are correct, the teacher's mobile terminal obtains the real-time video screen of the corresponding student laboratory table through the network, and manually determines whether the student's experimental step operation is passed or not.

[0004] In summary, the existing teaching system that uses AI target detection technology + specific experiments does not have the function of autonomous scoring. To address the above problems, the existing system needs to be improved. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent scoring system for making a model experiment of gas exchange between human lungs and the outside world, so as to solve the problem that the existing teaching system using AI target detection technology + specific experiments proposed in the above background technology does not have the function of autonomous scoring.

[0006] To achieve the above-mentioned purpose, the present invention provides the following technical solution: an intelligent scoring system for making a model experiment of gas exchange between human lungs and the outside world, comprising:

[0007] The hardware module consists of two parts: a student tablet with a camera and a server with high computing power. The student tablet with a camera consists of a camera unit, an interactive touch input unit and a display unit, while the server consists of an image decoding unit, a deep network unit, a logic parsing unit and an image encoding unit.

[0008] Preferably, after receiving the start signal of the human lung model experiment, the flat panel part transmits the start signal to the server side, and the camera unit starts working at the same time.

[0009] Preferably, after the camera unit starts working, the algorithm deployed on the server side is started, the real-time video stream of the human lung model experiment transmitted back by the camera unit is received, and the video stream is decoded.

[0010] Preferably, the human lung model is trained by a YOLO deep network to obtain a deep detection model specifically suitable for human lung model experiments.

[0011] Preferably, the decoded image information is received, and a series of black box operations are performed on the image using a trained deep detection network to output the position information of each target required by a subsequent algorithm.

[0012] Preferably, the targets required for the output subsequent algorithm include a glass cover, a rubber stopper, a rubber membrane and a balloon.

[0013] Preferably, the scoring points of the algorithm include a. two balloons are respectively put on the ends of the Y-shaped tube; b. the Y-shaped tube connected with the balloons is connected to the rubber plug; c. the assembled rubber plug is installed in the glass cover; d. the rubber membrane is fixed to the bottom of the glass cover; e. the process of inhalation is demonstrated; f. the process of exhalation is demonstrated; g. the options on the answer sheet are correct; h. the equipment is organized.

[0014] Preferably, the server part obtains the location information of each relevant target of the scoring point from the list of all target information output by the detection network according to different scoring points, and performs a series of logical judgments from the two dimensions of time and space to confirm whether the student's human lung model experimental operation complies with the specifications. If it complies with the operating specifications, the scoring point is assigned 1 point, and the camera screen of the time period is screenshoted as the basis for scoring to facilitate subsequent review, otherwise the scoring point is assigned 0 points.

[0015] Preferably, each time the depth detection model processes a frame of the image, the score will be transmitted back to the tablet, and the tablet will display the student's operation screen under the camera and the score of each scoring point in real time.

[0016] Preferably, the information of the student's interactive answering with the tablet is directly compared by the tablet, and the relevant scoring points are assigned points. The display of the scoring situation of the scoring points will be updated in real time until the end of the test.

[0017] Compared with the prior art, the invention has the following beneficial effects: the intelligent scoring system for making a model experiment of human lung and external gas exchange is

[0018] The present invention can effectively solve the problem that the existing teaching system using AI target detection technology + specific experiments does not have the function of autonomous scoring by using a student tablet with a camera and a server with high computing power. YOLO is small and fast, and the business logic code for the human lung model experiment takes a very short time. The present invention can directly give the score of the experimental operation in real time while the students are doing the experiment, which can be used for test scoring or teaching assistance. Due to the uniqueness of the present invention and the human lung model experiment as a part of the physical, chemical and biological experiment, the emergence of the present invention also improves the integrity of the intelligent scoring of the physical, chemical and biological experiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a system framework diagram of the present invention;

[0020] Figure 2 The figure is a flow chart of the algorithm of the present invention. DETAILED DESCRIPTION

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] See also Figure 1-2 The present invention provides a technical solution: an intelligent scoring system for making a model experiment of gas exchange between human lungs and the outside world, comprising:

[0023] Hardware module, the hardware module consists of two parts: a student tablet with a camera and a server with high computing power. The student tablet with a camera is composed of a camera unit, an interactive touch input unit and a display unit. At the same time, the server is composed of an image decoding unit, a deep network unit, a logic analysis unit and an image encoding unit.

[0024] After receiving the start signal of the human lung model experiment, the tablet part transmits the start signal to the server, and the camera unit starts working at the same time.

[0025] After the camera unit starts working, the algorithm deployed on the server side is started, which receives the real-time video stream of the human lung model experiment sent back by the camera unit and decodes the video stream.

[0026] The human lung model is trained with the YOLO deep network to obtain a deep detection model specifically suitable for human lung model experiments.

[0027] After receiving the decoded image information, the trained deep detection network is used to perform a series of black box operations on the image, and the position information of each target required by the subsequent algorithm is output.

[0028] The targets required for the subsequent output algorithms include glass covers, rubber stoppers, rubber membranes, and balloons.

[0029] The scoring points for the algorithm include: a. putting two balloons on the ends of the Y-shaped tube respectively; b. connecting the Y-shaped tube with the balloons and the rubber plug; c. putting the assembled rubber plug into the glass cover; d. fixing the rubber membrane at the bottom of the glass cover; e. demonstrating the process of inhalation; f. demonstrating the process of exhalation; g. ensuring that the options on the answer sheet are correct; h. organizing the equipment.

[0030] The server part obtains the location information of each relevant target of the scoring point from the list of all target information output by the detection network according to the different scoring points, and performs a series of logical judgments from the two dimensions of time and space to confirm whether the student's human lung model experimental operation complies with the specifications. If it complies with the operating specifications, the scoring point is assigned 1 point, and the camera screen of the time period is screenshoted as the basis for scoring to facilitate subsequent review. Otherwise, the scoring point is assigned 0 points.

[0031] Every time the depth detection model processes a frame of the image, it will send the score back to the tablet, which will then display the student's operation screen under the camera and the score of each scoring point in real time.

[0032] The information of students' interactive answers with the tablet will be directly compared with the data on the tablet, and the relevant scoring points will be assigned points. The display of the scoring points will be updated in real time until the end of the exam.

[0033] The present invention is based on a student tablet with a camera at a specific angle and a server with high computing power, and includes the following steps:

[0034] S1: Read the real-time video stream of the camera, or the student experiment operation video recorded by the camera.

[0035] S2: Pass the video stream / video frame by frame into the YOLO detection network for target detection operation.

[0036] S3: Perform logical judgment on the target information output by the deep network to obtain the score information of each score point.

[0037] S4: The score information is transmitted back to the student tablet.

[0038] S5: The tablet compares the option information filled in by the student with the preset answers to determine whether the score point is scored.

[0039] S6: The student tablet displays the camera image, score and experiment prompts in real time.

[0040] Specifically, the present invention is unique. Currently, there is no AI scoring system for physical, chemical and biological experiments on the market. In terms of specific experimental applications, our system is unique. Secondly, the present invention is real-time. Since YOLO is small and fast, and the business logic code for human lung model experiments takes a very short time, the present invention can directly give the scores of experimental operations in real time while students are doing experiments, which can be used for test grading or teaching assistance. Due to the uniqueness of the present invention and the human lung model experiment as a part of physical, chemical and biological experiments, the emergence of the present invention also improves the integrity of intelligent scoring of physical, chemical and biological experiments.

[0041] like Figure 2 As shown, in the actual process, students determine to make a model of gas exchange between the human lung and the outside world, namely, a human lung model. At the beginning of the experiment, after the tablet receives the start signal of the human lung model experiment, it transmits the start signal to the server, and at the same time the camera starts working, the algorithm deployed on the server starts, and the real-time video stream of the human lung model experiment sent back by the camera is received, and the video stream is decoded. In the early stage, a large amount of image data of the human lung model experiment is trained through the YOLO deep network to obtain a deep detection model specifically suitable for the human lung model experiment. After receiving the decoded image information, the trained deep detection network is used to perform a series of black box operations on the image, and the position information of each target glass cover, rubber stopper, rubber membrane, balloon, etc. required by the subsequent algorithm is output;

[0042] Then, according to the different scoring points, the location information of each relevant target of the scoring point is obtained from the list of all target information output by the detection network. A series of logical judgments are made from the two dimensions of time and space to confirm whether the student's human lung model experimental operation meets the specifications. If it meets the operating specifications, the scoring point is assigned 1 point, and the camera screen of the period is screenshoted as the basis for scoring to facilitate subsequent review. Otherwise, the scoring point is assigned 0 points.

[0043] Each time a frame of image is processed, the score will be sent back to the tablet, which will display the student's operation screen under the camera and the score of each scoring point in real time. The tablet will directly compare the data of the student's interactive answering information with the tablet and assign points to the relevant scoring points. The display of the scoring point will be updated in real time until the end of the test.

[0044] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world, characterized in that: include: The hardware module consists of two parts: a student tablet with a camera and a server with high computing power. The student tablet with a camera consists of a camera unit, an interactive touch input unit and a display unit, while the server consists of an image decoding unit, a deep network unit, a logic parsing unit and an image encoding unit.

2. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 1, characterized in that: After receiving the start signal of the human lung model experiment, the flat panel part transmits the start signal to the server side, and the camera unit starts working at the same time.

3. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 1, characterized in that: After the camera unit starts working, the algorithm deployed on the server side is started, the real-time video stream of the human lung model experiment transmitted back by the camera unit is received, and the video stream is decoded.

4. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 3, characterized in that: The human lung model is trained by the YOLO deep network to obtain a deep detection model specifically suitable for human lung model experiments.

5. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 3 is characterized in that: After receiving the decoded image information, a series of black box operations are performed on the image using the trained deep detection network to output the position information of each target required by the subsequent algorithm.

6. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 5, characterized in that: The targets required by the output subsequent algorithm include a glass cover, a rubber stopper, a rubber membrane and a balloon.

7. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 6, characterized in that: The scoring points of the algorithm include a. putting two balloons on the ends of the Y-shaped tube respectively; b. connecting the Y-shaped tube with the balloons and the rubber plug; c. putting the assembled rubber plug into the glass cover; d. fixing the rubber membrane at the bottom of the glass cover; e. demonstrating the process of inhalation; f. demonstrating the process of exhalation; g. correct options on the answer sheet; h. organizing the equipment.

8. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 1, characterized in that: The server part obtains the location information of each relevant target of the scoring point from the list of all target information output by the detection network according to different scoring points, and performs a series of logical judgments from the two dimensions of time and space to confirm whether the student's human lung model experimental operation complies with the specifications. If it complies with the operating specifications, the scoring point is assigned 1 point, and the camera screen of the time period is screenshoted as the basis for scoring to facilitate subsequent review. Otherwise, the scoring point is assigned 0 points.

9. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 4, characterized in that: Each time the depth detection model processes a frame of the image, it will transmit the score back to the tablet, which will display the student's operation screen under the camera and the score of each scoring point in real time.

10. The intelligent scoring system for making a model experiment of gas exchange between human lung and the outside world according to claim 1, characterized in that: The tablet computer will directly compare the information of the students' interactive answers with the tablet computer, and assign points to the relevant scoring points. The display of the scoring situation of the scoring points will be updated in real time until the end of the test.

Images