Man-machine interaction system

Abstract

The invention relates to a man-machine interaction system which comprises a content learning mechanism, a dynamic shooting mechanism, a mountain analysis mechanism, a data detection mechanism, a state judgment mechanism and a man-machine interaction mechanism. The man-machine interaction system is reliable in logic and easy and convenient to control.

Description

Technical Field

[0001] The present invention relates to the field of human-computer interaction, and in particular to a human-computer interaction system. Background Art

[0002] The future development direction of human-computer interaction is multifaceted, covering multiple levels such as technological innovation, application expansion and theoretical research. The following are some key development trends:

[0003] 1. Brain-computer interface technology: Brain-computer interface (BCI) technology allows neural signals to be directly obtained from the brain to control external devices, which is particularly significant in the field of medical health. BCI technology allows users to control external devices with just their thoughts, such as providing motor recovery solutions for paralyzed patients or achieving unprecedented immersive experience in games. However, the accompanying ethical and privacy issues also need to be resolved. How to find a balance between technological progress and personal rights will be one of the focuses of future research.

[0004] 2. Augmented reality (AR) and virtual reality (VR): These technologies will provide a more immersive and natural interactive experience and may become an important platform for future human-computer interaction. Mixed reality (MR) will bring revolutionary changes to education, training, entertainment and other fields. Through ultra-realistic interactive scenes, users can freely shuttle between the virtual and real worlds and experience a new realm of learning and entertainment. For example, in history classes, students can "experience" ancient battlefields and feel the shock of historical events, which greatly enhances the fun and effectiveness of learning.

[0005] 3. Smart cars and smart transportation: The development of smart driving and smart transportation systems will further promote the application of human-computer interaction technology in the automotive field.

[0006] CN119336959A discloses a human-computer interaction method and electronic device, which relate to the field of AI. The method is applied to an electronic device, including: receiving a user instruction, and in a scenario where the user instruction represents a certain scene intention, obtaining multiple target APIs according to the scene intention, and the multiple target APIs correspond to the scene intention. Then output the results of calling the multiple target APIs, and the results of calling the multiple APIs can be used to realize the user needs corresponding to the scene intention. In this way, based on the user's scene intention, the electronic device can directly output the results of calling the API that meet the user needs under the scene intention, and the API is a direct exposure of the application function. By directly outputting the results of calling multiple APIs, a richer, more comprehensive and convenient response result is achieved, which can meet user needs and improve the effect of human-computer dialogue.

[0007] CN119336154A relates to the field of human-computer interaction, and discloses a human-computer interaction method, device, computer equipment and storage medium. The technology collects first positioning information obtained by a hand-worn device and hand image information collected by an eye-worn device, detects second positioning information in the hand image information, and can accurately determine the position information of the hand in space based on pre-acquired parameter information of a camera device and hand image information; the first positioning information can be supplemented based on the second positioning information and spatial position information to obtain accurate hand input information, thereby improving the accuracy of three-dimensional input information and reducing the power consumption of wearable devices. Summary of the invention

[0008] In order to overcome the technical problems in the prior art, the present invention proposes a human-computer interaction system, which can perform multiple learning operations on a convolutional neural network to obtain a convolutional neural network after completing multiple learning operations, and output the convolutional neural network after completing multiple learning operations. The number of learning operations completed by the convolutional neural network is inversely correlated with the shooting frame rate of the dynamic shooting mechanism, so that convolutional neural networks with different structures are set for different dynamic shooting mechanisms, and when there is a mine in the preview picture, the shooting action of the shooting frame rate of the detection environment in front of the mine detection robot is triggered to obtain each frame of the detection environment picture corresponding to each time stamp, obtain the image block where the mountain is located in each received frame of the detection environment picture as the target image block of the frame of the detection environment picture, identify the horizontal coordinate values, vertical coordinate values ​​and depth of field values ​​corresponding to each edge pixel point of the target image block, and calculate the total number of each edge pixel point of the target image block, the horizontal coordinate values, vertical coordinate values ​​and depth of field values ​​corresponding to each edge pixel point of the target image block. The deep value is used as the visualization information corresponding to the target image block, and the shooting frame rate, imaging focal length, exposure and aperture value of the dynamic shooting mechanism are obtained as the various shooting data outputs of the dynamic shooting mechanism, so as to provide sufficient and comprehensive basic information for subsequent intelligent judgment. The convolutional neural network after completing multiple learning operations is also used to intelligently judge whether a collapse occurs in the mine ahead within the next time segment of the preset time segment based on the visualization information corresponding to the target image block of the multi-frame detection environment picture corresponding to multiple timestamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment. A human-computer interaction mechanism is also used to send an emergency avoidance request to the operating user of the mine detection robot when the received collapse prediction mark indicates that the mine ahead will collapse within the next time segment of the preset time segment. Otherwise, an environmental safety signal is sent to the operating user of the mine detection robot. Therefore, on the basis of effectively predicting whether there is a risk of collapse in the mine ahead in the future, the human-computer interaction mechanism is used to ensure the safety of the mine detection robot.

[0009] According to the present invention, a human-computer interaction system is provided, the system comprising: A content learning mechanism is provided in the mine detection robot, and is used to perform multiple learning operations on the convolutional neural network to obtain a convolutional neural network after completing the multiple learning operations, and output the convolutional neural network after completing the multiple learning operations, wherein the number of learning operations completed by the convolutional neural network is inversely correlated with the shooting frame rate of the dynamic shooting mechanism; The dynamic shooting mechanism is arranged in the front screen of the mine detection robot, and is used to trigger the shooting action of the shooting frame rate of the detection environment in front of the mine detection robot when there is a mine in the preview picture, so as to obtain each frame of the detection environment picture corresponding to each time stamp; A mountain analysis mechanism is arranged in the mine detection robot and connected to the dynamic shooting mechanism, and is used to obtain the image block where the mountain is located in each received frame of the detection environment picture as the target image block of the detection environment picture, identify the respective horizontal coordinate values, the respective vertical coordinate values, and the respective depth of field values ​​corresponding to the respective edge pixel points of the target image block, and use the total number of the respective edge pixel points of the target image block, the respective horizontal coordinate values, the respective vertical coordinate values, and the respective depth of field values ​​corresponding to the respective edge pixel points of the target image block as the visualization information corresponding to the target image block; A data detection mechanism, connected to the dynamic shooting mechanism, disposed in the mine detection robot, and used to obtain the shooting frame rate, imaging focal length, exposure and aperture value of the dynamic shooting mechanism as various shooting data outputs of the dynamic shooting mechanism; A state judgment mechanism is arranged in the mine detection robot and is respectively connected to the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and is used to use a convolutional neural network after completing multiple learning operations to intelligently judge whether a collapse prediction mark will appear in the mine ahead in the next time segment of the preset time segment based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple timestamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment; A human-machine interaction mechanism is arranged in the mine detection robot and connected to the state judgment mechanism, and is used to send an emergency avoidance request to the operator of the mine detection robot when the received collapse prediction mark indicates that the mine ahead will collapse within the next time segment of the preset time segment, and to send an environmental safety signal to the operator of the mine detection robot when the received collapse prediction mark indicates that the mine ahead will not collapse within the next time segment of the preset time segment; Among them, the collapse prediction mark indicating whether the mine ahead will collapse in the next time segment of the preset time segment is intelligently judged by using a convolutional neural network after completing multiple learning operations based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple time stamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment, including: using two collapse prediction marks with different values ​​to respectively indicate whether the mine ahead will collapse in the next time segment of the preset time segment; Among them, a convolutional neural network that has completed multiple learning operations is used to intelligently judge whether a collapse will occur in the mine ahead within the next time segment of the preset time segment based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple timestamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment. The collapse prediction mark also includes: the next time segment of the preset time segment and the preset time segment have equal duration lengths.

[0010] It can be seen that the present invention has at least the following important inventive concepts: Inventive concept A: performing multiple learning operations on a convolutional neural network to obtain a convolutional neural network after completing the multiple learning operations, and outputting the convolutional neural network after completing the multiple learning operations, wherein the number of learning operations completed by the convolutional neural network is inversely related to the shooting frame rate of the dynamic shooting mechanism, thereby setting convolutional neural networks with different structures for different dynamic shooting mechanisms; Invention concept B: When there is a mine in the preview image, trigger the shooting action of the shooting frame rate of the detection environment in front of the mine detection robot to obtain each frame of the detection environment image corresponding to each time stamp, obtain the image block where the mountain is located in each received frame of the detection environment image as the target image block of the detection environment image, identify the horizontal coordinate values, vertical coordinate values ​​and depth of field values ​​corresponding to each edge pixel point of the target image block, and use the total number of each edge pixel point of the target image block, the horizontal coordinate values, vertical coordinate values ​​and depth of field values ​​corresponding to each edge pixel point of the target image block as the visualization information corresponding to the target image block, and at the same time obtain the shooting frame rate, imaging focal length, exposure and aperture value of the dynamic shooting mechanism as the shooting data output of the dynamic shooting mechanism, so as to provide sufficient and comprehensive basic information for subsequent intelligent judgment; Inventive concept C: A convolutional neural network that has completed multiple learning operations is used to intelligently judge a collapse prediction mark indicating whether a mine ahead will collapse within the next time segment of the preset time segment based on visualization information corresponding to target image blocks of multiple frames of detection environment pictures corresponding to multiple time stamps within a preset time segment, various shooting data of a dynamic shooting mechanism, and the duration length of the preset time segment. A human-computer interaction mechanism is also used to send an emergency avoidance request to the operating user of the mine detection robot when the received collapse prediction mark indicates that a mine ahead will collapse within the next time segment of the preset time segment. Otherwise, an environmental safety signal is sent to the operating user of the mine detection robot. Therefore, on the basis of effectively predicting whether there is a risk of collapse in the mine ahead in the future, the human-computer interaction mechanism is used to ensure the safety of the mine detection robot.

[0011] The human-machine interaction system of the present invention has reliable logic and is easy to operate. Since it can intelligently judge whether a collapse prediction mark indicates whether a collapse occurs in the mine ahead within the next time segment of the preset time segment, a human-machine interaction mechanism is also used to send an emergency avoidance request to the operator of the mine detection robot when the received collapse prediction mark indicates that a collapse occurs, thereby using a human-machine interaction mechanism to ensure the safety of the mine detection robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] The embodiments of the present invention will be described below with reference to the accompanying drawings, wherein:

[0013] Figure 1 FIG. 1 is a block diagram of the internal structure of a human-computer interaction system according to Embodiment 1 of the present invention.

[0014] Figure 2 FIG. 2 is a block diagram of the internal structure of a human-computer interaction system according to Embodiment 2 of the present invention.

[0015] Figure 3 1 is a block diagram of the internal structure of a human-computer interaction system according to Embodiment 3 of the present invention. DETAILED DESCRIPTION

[0016] In the prior art, detection in the field of danger detection can also adopt a human-computer interaction system under remote control, for example, the human-computer interaction system of a mine detection robot. It is hoped that the human-computer interaction mechanism of the mine detection robot can intelligently predict whether the mine ahead will collapse in the future time segment, and send environmental safety signals or environmental danger signals to the remote operating user of the mine detection robot based on the intelligent prediction results, so as to improve the intelligence level and safety of mine detection. Obviously, there is a lack of mature solutions in the prior art.

[0017] The implementation scheme of the human-computer interaction system of the present invention will be described in detail below with reference to the accompanying drawings.

[0018] Figure 1 The internal structure block diagram of the human-computer interaction system according to Embodiment 1 of the present invention is shown, and the system includes: A content learning mechanism is provided in the mine detection robot, and is used to perform multiple learning operations on the convolutional neural network to obtain a convolutional neural network after completing the multiple learning operations, and output the convolutional neural network after completing the multiple learning operations, wherein the number of learning operations completed by the convolutional neural network is inversely correlated with the shooting frame rate of the dynamic shooting mechanism; Specifically, the content learning mechanism is arranged in the mine detection robot, and is used to perform multiple learning operations on the convolutional neural network to obtain the convolutional neural network after completing the multiple learning operations, and output the convolutional neural network after completing the multiple learning operations, and the number of learning operations completed by the convolutional neural network is inversely correlated with the shooting frame rate of the dynamic shooting mechanism, including: a numerical mapping function can be used to represent the numerical mapping relationship of the number of learning operations completed by the convolutional neural network and the shooting frame rate of the dynamic shooting mechanism. The dynamic shooting mechanism is arranged in the front screen of the mine detection robot, and is used to trigger the shooting action of the shooting frame rate of the detection environment in front of the mine detection robot when there is a mine in the preview picture, so as to obtain each frame of the detection environment picture corresponding to each time stamp; A mountain analysis mechanism is arranged in the mine detection robot and connected to the dynamic shooting mechanism, and is used to obtain the image block where the mountain is located in each received frame of the detection environment picture as the target image block of the detection environment picture, identify the respective horizontal coordinate values, the respective vertical coordinate values, and the respective depth of field values ​​corresponding to the respective edge pixel points of the target image block, and use the total number of the respective edge pixel points of the target image block, the respective horizontal coordinate values, the respective vertical coordinate values, and the respective depth of field values ​​corresponding to the respective edge pixel points of the target image block as the visualization information corresponding to the target image block; A data detection mechanism, connected to the dynamic shooting mechanism, disposed in the mine detection robot, and used to obtain the shooting frame rate, imaging focal length, exposure and aperture value of the dynamic shooting mechanism as various shooting data outputs of the dynamic shooting mechanism; A state judgment mechanism is arranged in the mine detection robot and is respectively connected to the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and is used to use a convolutional neural network after completing multiple learning operations to intelligently judge whether a collapse prediction mark will appear in the mine ahead in the next time segment of the preset time segment based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple timestamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment; A human-machine interaction mechanism is arranged in the mine detection robot and connected to the state judgment mechanism, and is used to send an emergency avoidance request to the operator of the mine detection robot when the received collapse prediction mark indicates that the mine ahead will collapse within the next time segment of the preset time segment, and to send an environmental safety signal to the operator of the mine detection robot when the received collapse prediction mark indicates that the mine ahead will not collapse within the next time segment of the preset time segment; Among them, the collapse prediction mark indicating whether the mine ahead will collapse in the next time segment of the preset time segment is intelligently judged by using a convolutional neural network after completing multiple learning operations based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple time stamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment, including: using two collapse prediction marks with different values ​​to respectively indicate whether the mine ahead will collapse in the next time segment of the preset time segment; Among them, the collapse prediction mark indicating whether the mine ahead will collapse in the next time segment of the preset time segment is intelligently judged by using the convolutional neural network after completing multiple learning operations based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple time stamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment, and further includes: the duration lengths of the next time segment of the preset time segment and the preset time segment are equal; Among them, the content learning mechanism is arranged in the mining detection robot, and is used to perform multiple learning operations on the convolutional neural network to obtain the convolutional neural network after completing the multiple learning operations, and output the convolutional neural network after completing the multiple learning operations. The number of learning operations completed by the convolutional neural network is inversely correlated with the shooting frame rate of the dynamic shooting mechanism, including: the larger the value of the shooting frame rate of the dynamic shooting mechanism, the fewer the number of learning operations completed by the convolutional neural network.

[0019] Figure 2 FIG. 2 is a block diagram of the internal structure of a human-computer interaction system according to Embodiment 2 of the present invention.

[0020] Compared to Figure 1 , Figure 2 The human-computer interaction system in can also include the following components: The dust measuring mechanism comprises a plurality of dust measuring units, which are used to respectively measure the current surface dust concentration values ​​of the state judging mechanism, the data detecting mechanism, the mountain analyzing mechanism and the content learning mechanism; Wherein, the dust measuring mechanism includes a plurality of dust measuring units, which are used to respectively measure the current surface dust concentration values ​​of the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, including: the plurality of dust measuring units respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism are a plurality of non-contact dust concentration sensors; Wherein, the dust measuring mechanism includes a plurality of dust measuring units for respectively measuring the current surface dust concentration values ​​of the state judging mechanism, the data detecting mechanism, the mountain analyzing mechanism and the content learning mechanism, and further includes: the internal structures of a plurality of non-contact dust concentration sensors respectively used by the state judging mechanism, the data detecting mechanism, the mountain analyzing mechanism and the content learning mechanism are the same; Wherein, the dust measurement mechanism includes a plurality of dust measurement units for respectively measuring the current surface dust concentration values ​​of the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and further includes: a plurality of non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism have the same dust measurement upper limit threshold and dust measurement lower limit threshold; Among them, the dust measuring mechanism includes multiple dust measuring units, which are used to respectively measure the current surface dust concentration values ​​of the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism. It also includes: the distances between the multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism are equal.

[0021] Figure 3 1 is a block diagram of the internal structure of a human-computer interaction system according to Embodiment 3 of the present invention.

[0022] Compared to Figure 1 , Figure 3 The human-computer interaction system in can also include the following components: a content notification device, connected to the plurality of non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and used to perform corresponding dust alarm actions based on dust concentration measurement results of the plurality of non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism; The content notification device is respectively connected to a plurality of non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and is used to perform corresponding dust alarm actions based on dust concentration measurement results of the plurality of non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, including: the content notification device has a built-in dust receiving unit, a dust judgment unit and a notification execution unit; And wherein, the content notification device is respectively connected to multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and is used to perform corresponding dust alarm actions based on the dust concentration measurement results of the multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and also includes: within the content notification device, the dust receiving unit, the dust judgment unit and the notification execution unit are connected in sequence.

[0023] In addition, in the human-computer interaction system, a convolutional neural network that has completed multiple learning operations is used to intelligently judge whether a collapse will occur in the mine ahead within the next time segment of the preset time segment based on the visualization information corresponding to the target image blocks of the multiple frames of detection environment pictures corresponding to multiple timestamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment. The method also includes: synchronously inputting the visualization information corresponding to the target image blocks of the multiple frames of detection environment pictures corresponding to multiple timestamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment into the convolutional neural network that has completed multiple learning operations, and executing the convolutional neural network that has completed multiple learning operations to obtain a collapse prediction mark that is output by the convolutional neural network that has completed multiple learning operations, indicating whether a collapse will occur in the mine ahead within the next time segment of the preset time segment.

[0024] Although the present invention has been described with reference to the preferred embodiments, it is obvious that various modifications and variations can be made to the present invention by those skilled in the art without departing from the spirit and scope of the present invention. Therefore, various modifications and variations of the present invention are covered by the appended claims and their equivalents.

Claims

1. A human-computer interaction system, characterized in that: The system comprises: A content learning mechanism is provided in the mine detection robot, and is used to perform multiple learning operations on the convolutional neural network to obtain a convolutional neural network after completing the multiple learning operations, and output the convolutional neural network after completing the multiple learning operations, wherein the number of learning operations completed by the convolutional neural network is inversely correlated with the shooting frame rate of the dynamic shooting mechanism; The dynamic shooting mechanism is arranged in the front screen of the mine detection robot, and is used to trigger the shooting action of the shooting frame rate of the detection environment in front of the mine detection robot when there is a mine in the preview picture, so as to obtain each frame of the detection environment picture corresponding to each time stamp; A mountain analysis mechanism is arranged in the mine detection robot and connected to the dynamic shooting mechanism, and is used to obtain the image block where the mountain is located in each received frame of the detection environment picture as the target image block of the detection environment picture, identify the respective horizontal coordinate values, the respective vertical coordinate values, and the respective depth of field values ​​corresponding to the respective edge pixel points of the target image block, and use the total number of the respective edge pixel points of the target image block, the respective horizontal coordinate values, the respective vertical coordinate values, and the respective depth of field values ​​corresponding to the respective edge pixel points of the target image block as the visualization information corresponding to the target image block; A data detection mechanism, connected to the dynamic shooting mechanism, disposed in the mine detection robot, and used to obtain the shooting frame rate, imaging focal length, exposure and aperture value of the dynamic shooting mechanism as various shooting data outputs of the dynamic shooting mechanism; A state judgment mechanism is arranged in the mine detection robot and is respectively connected to the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and is used to use a convolutional neural network after completing multiple learning operations to intelligently judge whether a collapse prediction mark will appear in the mine ahead in the next time segment of the preset time segment based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple timestamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment; A human-machine interaction mechanism is arranged in the mine detection robot and connected to the state judgment mechanism, and is used to send an emergency avoidance request to the operator of the mine detection robot when the received collapse prediction mark indicates that the mine ahead will collapse within the next time segment of the preset time segment, and to send an environmental safety signal to the operator of the mine detection robot when the received collapse prediction mark indicates that the mine ahead will not collapse within the next time segment of the preset time segment; Among them, the collapse prediction mark indicating whether the mine ahead will collapse in the next time segment of the preset time segment is intelligently judged by using a convolutional neural network after completing multiple learning operations based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple time stamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment, including: using two collapse prediction marks with different values ​​to respectively indicate whether the mine ahead will collapse in the next time segment of the preset time segment; Among them, a convolutional neural network that has completed multiple learning operations is used to intelligently judge whether a collapse will occur in the mine ahead within the next time segment of the preset time segment based on the visualization information corresponding to the target image blocks of the multi-frame detection environment screen corresponding to multiple timestamps in the preset time segment, the various shooting data of the dynamic shooting mechanism and the duration length of the preset time segment. The collapse prediction mark also includes: the next time segment of the preset time segment and the preset time segment have equal duration lengths.

2. The human-computer interaction system according to claim 1, characterized in that: A content learning mechanism is arranged in the mining detection robot, and is used to perform multiple learning operations on the convolutional neural network to obtain the convolutional neural network after completing the multiple learning operations, and output the convolutional neural network after completing the multiple learning operations. The number of learning operations completed by the convolutional neural network is inversely correlated with the shooting frame rate of the dynamic shooting mechanism, including: the larger the shooting frame rate of the dynamic shooting mechanism, the smaller the number of learning operations completed by the convolutional neural network.

3. The human-computer interaction system according to claim 2, characterized in that: The system further comprises: The dust measuring mechanism comprises a plurality of dust measuring units, which are used to respectively measure the current surface dust concentration values ​​of the state judging mechanism, the data detecting mechanism, the mountain analyzing mechanism and the content learning mechanism; Among them, the dust measuring mechanism includes multiple dust measuring units, which are used to respectively measure the current surface dust concentration values ​​of the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, including: the multiple dust measuring units respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism are multiple non-contact dust concentration sensors.

4. The human-computer interaction system according to claim 3, characterized in that: The dust measuring mechanism includes multiple dust measuring units, which are used to respectively measure the current surface dust concentration values ​​of the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism. It also includes: the internal structures of multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism are the same.

5. The human-computer interaction system according to claim 3, characterized in that: The dust measuring mechanism includes multiple dust measuring units, which are used to respectively measure the current surface dust concentration values ​​of the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism. It also includes: multiple non-contact dust concentration sensors respectively used for the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism have the same dust measurement upper limit threshold and dust measurement lower limit threshold.

6. The human-computer interaction system according to claim 5, characterized in that: The dust measuring mechanism includes multiple dust measuring units, which are used to respectively measure the current surface dust concentration values ​​of the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism. It also includes: the distances between the multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism are equal.

7. The human-computer interaction system according to claim 3, characterized in that: The system further comprises: The content notification device is respectively connected to multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and is used to perform corresponding dust alarm actions based on the dust concentration measurement results of the multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism.

8. The human-computer interaction system according to claim 7, characterized in that: The content notification device is respectively connected to multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and is used to perform corresponding dust alarm actions based on the dust concentration measurement results of the multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, including: the content notification device has a built-in dust receiving unit, a dust judgment unit and a notification execution unit.

9. The human-computer interaction system according to claim 8, characterized in that: The content notification device is respectively connected to multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism, and is used to perform corresponding dust alarm actions based on the dust concentration measurement results of the multiple non-contact dust concentration sensors respectively used by the state judgment mechanism, the data detection mechanism, the mountain analysis mechanism and the content learning mechanism. It also includes: within the content notification device, the dust receiving unit, the dust judgment unit and the notification execution unit are connected in sequence.

Images