Detection robot control system applied to emergencies

Abstract

The invention discloses a detection robot control system applied to emergencies, and relates to the technical field of emergency rescue, the system discloses a physical sign detection module, a rescue judgment module and a supplementary rescue module, by arranging the rescue judgment module and the physical sign detection module, a detection robot can be controlled to recognize vital signs in a rescue area, and the rescue judgment module is controlled to supplement the rescue area. According to a detection video shot by a detection robot, whether the detection video needs to be uploaded or not is judged, the detection video is judged to be sent to a rescue expert group or a common rescue group, the rescue video is distributed intelligently, the rescue group does not need to check the rescue video any more, the rescue efficiency of emergencies is improved, a supplementary rescue module is arranged, and the rescue efficiency is improved. The fixed assistance robot can be controlled to move to the expert rescue position to shoot videos, multi-angle video shooting is controlled to be provided for personnel needing expert rescue of the detection robot, and the rescue expert is assisted to conduct multi-angle rescue work.

Description

Technical Field

[0001] The present invention relates to the technical field of emergency rescue, and more specifically, it relates to a detective robot control system applied to emergencies. Background Art

[0002] An emergency refers to a natural disaster, an accident disaster, a public health event, and a social security event that occur suddenly and cause or may cause serious social harm and require emergency response measures to deal with. Especially when encountering emergencies such as earthquakes and collapses, only detective robots can move in narrow gaps, and the detective robots upload the captured videos to the rescue team. When the rescue team receives the videos, rescue personnel need to first judge the vital sign status of the people in the videos through the videos and then judge whether rescue expert teams are needed for rescue guidance based on the vital sign status. However, this judgment process often easily delays the best rescue guidance time.

[0003] The prior art with the application number discloses. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the purpose of the present invention is to provide a detective robot control system applied to emergencies.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A detective robot control system applied to emergencies, including a vital sign detective module, a rescue determination module, and a supplementary rescue module;

[0007] The vital sign detective module is used to control the detective robot to move in the rescue area. While moving, the detective robot detects whether there are vital signs around. When it detects that there are vital signs around, the detective robot stops moving and takes a video in the direction where the vital signs exist, and marks the captured video as a detective video;

[0008] The rescue determination module is used to determine whether to upload the detective video and the target terminal for video upload according to video frames, specifically:

[0009] Convert the detective video into video frames, use the video frames as the input data of the physical sign detection model, obtain the output data of the physical sign detection model, mark the training label of the output data of the physical sign detection model as the physical sign detection value, sort all the physical sign detection values in the order of the corresponding video frames, calculate the difference between the next physical sign detection value and the previous one after sorting to obtain the physical sign change value. When the physical sign change value > 0, mark this physical sign change value as the physical sign false change value, obtain the total physical sign false value Fy, set the physical sign detection threshold. When the physical sign detection value ≥ the physical sign detection threshold, mark this physical sign detection value as the detection false value, obtain the weakness detection value Sc, obtain the detective warning value Lw of this detective video, set the high detective warning value as Mb, set the low detective warning value as Zj. When the detective warning value Lw ≥ the high detective warning value Mb, this detective robot immediately sends the detective video to the terminal of the rescue expert group, mark this detective robot as an expert rescue robot, and mark the position where this detective robot takes the video as the expert rescue position. When the low detective warning value Zj ≤ the detective warning value Lw < the high detective warning value Mb, this detective robot immediately sends the detective video to the terminal of the general rescue group. When the detective warning value Lw < the low detective warning value Zj, send the position where this detective robot takes the video to the terminal of the general rescue group, and control the detective robot to continue moving to detect the vital signs;

[0010] The supplementary rescue module is used to control the fixed rescue robot to move to the expert rescue position to take videos, specifically:

[0011] Take the expert rescue position of the expert rescue robot as the center, draw a circle with a preset radius to obtain the supplementary rescue range, mark the moving detective robots within the supplementary rescue range as the robots to be supplemented, obtain the supplementary rescue value Jq of the robots to be supplemented, mark the robot to be supplemented with the largest supplementary rescue value as the fixed rescue robot, control the fixed rescue robot to move to the expert rescue position, and the expert rescue robot and the fixed rescue robot take videos of the expert rescue position from multiple angles.

[0012] Further, the physical sign detection model is obtained through the following steps: Obtain multiple image frames, mark the image frames as training images, assign training labels to the training images, divide the training images into a training set and a validation set according to a set ratio, construct a neural network model, and perform iterative training on the neural network model through the training set and the validation set. When the number of iterative training times is greater than the iterative threshold, it is determined that the neural network model is trained, and mark the trained neural network model as the physical sign detection model. The larger the numerical value of the training label of the output data of the physical sign detection model, the weaker the vital signs in the image frame.

[0013] Further, the total virtual value Fy of the physical signs is obtained through the following steps: Mark the virtual change value of the physical signs as Ri, set the virtual change value coefficient of the physical signs as Pq, where q = 1, 2, 3, …, q; P1 < P2 < P3 < … < Pq. Set a range of virtual change values of the physical signs corresponding to each virtual change value coefficient, including (0, R1], (R1, R2], …, (Ri-1, Ri]. When Ri ∈ (0, R1], the corresponding virtual change value coefficient takes the value of P1. Using the formula obtain the total virtual change value Gt of the physical signs, where Lm is the total number of virtual change values of the physical signs. Use the formula Fy = Gt × a1 + Lm × a2 to obtain the total virtual value Fy of the physical signs, where a1 is the total virtual change value coefficient of the physical signs and a2 is the virtual change quantity coefficient of the physical signs.

[0014] Further, the weakness detection value Sc is obtained through the following steps: Calculate the difference between the detected virtual value and the physical sign detection threshold to obtain the detected virtual difference. Sum up all the detected virtual differences and take the average to obtain the average detected virtual difference, which is marked as Dz. Obtain the total number of times the physical sign detection value is marked as the detected virtual value, which is marked as Sg. Sort all the detected virtual values in the order of the corresponding video frames. Calculate the time difference between the times corresponding to two adjacent video frames after sorting to obtain the detected virtual value interval. Sum up all the detected virtual value intervals and take the average to obtain the average virtual value interval, which is marked as Pe. Using the formula obtain the weakness detection value Sc, where b1 is the average detected virtual difference coefficient, b2 is the detected virtual value number coefficient, and b3 is the average virtual value interval.

[0015] Further, the detection warning value Lw of the detective video is obtained through the following steps: Use the formula Lw = Fy × c1 + Sc × c2 to obtain the detection warning value Lw of the detective video, where c1 is the total virtual value coefficient of the physical signs and c2 is the weakness detection value coefficient.

[0016] Further, the supplementary rescue value Jq of the robot to be supplemented is obtained through the following steps: Obtain the position of the robot to be supplemented. Calculate the difference between the position of the robot to be supplemented and the expert rescue position to obtain the to-be-supplemented rescue distance, which is marked as Lk. Obtain the remaining power of the robot to be supplemented, which is marked as Ns. Using the formula obtain the supplementary rescue value Jq of the robot to be supplemented, where d1 is the to-be-supplemented rescue distance coefficient and d2 is the remaining power coefficient.

[0017] Further, when the physical sign change value ≤ 0, no processing is performed.

[0018] Further, when the physical sign detection value < the physical sign detection threshold, no processing is performed.

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] 1. By setting up a rescue determination module and a vital sign detection module, it can control the detection robot to identify vital signs in the rescue area, and determine whether to upload the detection video based on the detection video captured by the detection robot, as well as determine whether to send the detection video to the rescue expert group or the general rescue group, intelligently allocate the rescue video, and no longer require the rescue group to review the rescue video, thus improving the rescue efficiency of emergencies;

[0021] 2. By setting up a supplementary rescue module, it can control the fixed rescue robot to move to the expert rescue position to capture videos, and control the multi-angle video capture of the personnel in need of expert rescue by the detection robot, assisting the rescue experts in multi-angle rescue work. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is the principle block diagram of the rescue determination module of the present invention;

[0023] Figure 2 It is the principle block diagram of the supplementary rescue module of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0024] Embodiment 1

[0025] Refer to Figures 1 to 2 , a detection robot control system applied to emergencies, including a vital sign detection module, a rescue determination module, and a supplementary rescue module.

[0026] The vital sign detection module is used to control the detection robot to move in the rescue area. While the detection robot is moving, it detects whether there are vital signs around. When vital signs are detected around, the detection robot stops moving and captures a video in the direction where the vital signs exist, and marks the captured video as the detection video.

[0027] The rescue determination module is used to determine whether to upload the detection video and the target terminal for video upload according to the video frames. Specifically:

[0028] Convert the detective video into video frames, use the video frames as the input data of the vital sign detection model, obtain the output data of the vital sign detection model, mark the training label of the output data of the vital sign detection model as the vital sign detection value. The vital sign detection model is obtained through the following steps: Obtain multiple image frames, mark the image frames as training images, assign training labels to the training images, divide the training images into a training set and a validation set according to a set ratio, construct a neural network model, and perform iterative training on the neural network model through the training set and the validation set. When the number of iterative training times is greater than the iteration number threshold, it is determined that the neural network model is trained. Mark the trained neural network model as the vital sign detection model. The larger the training label value of the output data of the vital sign detection model, the weaker the vital sign in the image frame. Sort all the vital sign detection values in the order of the corresponding video frames, calculate the difference between the adjacent subsequent vital sign detection value and the previous one after sorting to obtain the vital sign change value. When the vital sign change value > 0, mark the vital sign change value as the vital sign weakening change value, and obtain the total vital sign weakening value Fy. When the vital sign change value ≤ 0, no processing is performed. The total vital sign weakening value Fy is obtained through the following steps: Mark the vital sign weakening change value as Ri, set the vital sign weakening change value coefficient as Pq, q = 1, 2, 3,..., q; P1 < P2 < P3 <... < Pq. Set a range of vital sign weakening change values corresponding to each vital sign weakening change value coefficient, including (0, R1], (R1, R2],..., (Ri - 1, Ri]. When Ri ∈ (0, R1], the corresponding vital sign weakening change value coefficient takes the value of P1, and use the formula Obtain the total vital sign weakening value Gt, where Lm is the total number of vital sign weakening change values. Use the formula Fy = Gt × a1 + Lm × a2 to obtain the total vital sign weakening value Fy, where a1 is the total vital sign weakening value coefficient and a2 is the vital sign weakening quantity coefficient. The value of a1 is 0.48, and the value of a2 is 0.39. Set the vital sign detection threshold. When the vital sign detection value ≥ the vital sign detection threshold, mark the vital sign detection value as the detection weakening value, and obtain the weakening detection value Sc. When the vital sign detection value < the vital sign detection threshold, no processing is performed. The weakening detection value Sc is obtained through the following steps: Calculate the difference between the detection weakening value and the vital sign detection threshold to obtain the detection weakening difference. Sum up all the detection weakening differences and take the average to obtain the average detection weakening difference, and mark it as Dz. Obtain the total number of times the vital sign detection value is marked as the detection weakening value, and mark it as Sg. Sort all the detection weakening values in the order of the corresponding video frames, calculate the time difference between the times corresponding to two adjacent video frames after sorting to obtain the detection weakening value interval. Sum up all the detection weakening value intervals and take the average to obtain the average weakening value interval, and mark it as Pe. Use the formula Obtain the weakness detection value Sc, where b1 is the average detection virtual difference coefficient, b2 is the detection virtual value times coefficient, b3 is the average virtual value interval, the value of b1 is 0.56, the value of b2 is 0.25, and the value of b3 is 0.83. Obtain the detective warning value Lw of the detective video. The detective warning value Lw of the detective video is obtained through the following steps: Use the formula Lw = Fy×c1 + Sc×c2 to obtain the detective warning value Lw of the detective video, where c1 is the total virtual value coefficient of physical signs, c2 is the weakness detection value coefficient, the value of c1 is 0.69, and the value of c2 is 0.57. Set the detective warning high value as Mb and the detective warning low value as Zj. When the detective warning value Lw ≥ detective warning high value Mb, the detective robot immediately sends the detective video to the terminal of the rescue expert group, marks the detective robot as an expert rescue robot, and marks the location where the detective robot shoots the video as the expert rescue location. When the detective warning low value Zj ≤ detective warning value Lw < detective warning high value Mb, the detective robot immediately sends the detective video to the terminal of the ordinary rescue group. When the detective warning value Lw < detective warning low value Zj, send the location where the detective robot shoots the video to the terminal of the ordinary rescue group, and control the detective robot to continue moving to detect vital signs. When the detective warning value Lw of detective robot a ≥ detective warning high value Mb, detective robot a immediately sends the detective video to the terminal of the rescue expert group, marks detective robot a as an expert rescue robot, and marks the location where detective robot a shoots the video as the expert rescue location. When the detective warning low value Zj ≤ detective warning value Lw < detective warning high value Mb of detective robot b, detective robot b immediately sends the detective video to the terminal of the ordinary rescue group. When the detective warning value Lw of detective robot c < detective warning low value Zj, send the location where detective robot c shoots the video to the terminal of the ordinary rescue group, and control detective robot c to continue moving to detect vital signs. Set up a rescue determination module and a physical sign detective module, which can control the detective robot to identify vital signs in the rescue area, and determine whether to upload the detective video according to the detective video taken by the detective robot, and determine to send the detective video to the rescue expert group or the ordinary rescue group, intelligently allocate the rescue video, and no longer require the rescue group to review the rescue video, improving the rescue efficiency of emergencies.

[0029] The supplementary rescue module is used to control the fixed rescue robot to move to the expert rescue location to shoot videos, specifically:

[0030] Taking the expert rescue position of the expert rescue robot as the center, draw a circle with a preset radius to obtain a supplementary rescue range. Mark the detective robots that are moving and whose positions are within the supplementary rescue range as robots to be supplemented, and obtain the supplementary rescue value Jq of the robots to be supplemented. The supplementary rescue value Jq of the robots to be supplemented is obtained through the following steps: Obtain the position of the robot to be supplemented, calculate the difference between the position of the robot to be supplemented and the expert rescue position to obtain the supplementary rescue distance, and mark it as Lk. Obtain the remaining power of the robot to be supplemented and mark it as Ns. Use the formula to obtain the supplementary rescue value Jq of the robot to be supplemented, where d1 is the supplementary rescue distance coefficient, d2 is the remaining power coefficient, the value of d1 is 0.84, and the value of d2 is 0.65. Mark the robot to be supplemented with the largest supplementary rescue value as the fixed rescue robot, and control the fixed rescue robot to move to the expert rescue position. The expert rescue robot and the fixed rescue robot take videos of the expert rescue position from multiple angles. Set up a supplementary rescue module, which can control the fixed rescue robot to move to the expert rescue position to take videos, control the provision of multi-angle video shooting for the personnel in need of expert rescue by the detective robots, and assist the rescue experts in carrying out multi-angle rescue work.

[0031] Working principle:

[0032] Set up a rescue determination module and a physical sign detection module, which can control the detective robots to identify the vital signs in the rescue area, and determine whether to upload the detective videos according to the detective videos taken by the detective robots, and determine whether to send the detective videos to the rescue expert group or the ordinary rescue group, so as to intelligently allocate the rescue videos and no longer require the rescue group to review the rescue videos, improving the rescue efficiency of emergencies. Set up a supplementary rescue module, which can control the fixed rescue robot to move to the expert rescue position to take videos, control the provision of multi-angle video shooting for the personnel in need of expert rescue by the detective robots, and assist the rescue experts in carrying out multi-angle rescue work.

[0033] The above is only the preferred implementation mode of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as the protection scope of this template.

[0034] The above has described a detailed description of an embodiment of the present invention, but the content described is only the preferred embodiment of the present invention and cannot be considered as used to limit the implementation scope of the present invention. All equivalent changes and improvements made according to the scope of the present invention application should still fall within the scope covered by the patent of the present invention.

Claims

1. A detective robot control system applied to emergencies, characterized in that, It includes a physical sign detective module, a rescue determination module, and a supplementary rescue module; The physical sign detective module is used to control the detective robot to move within the rescue area. While moving, the detective robot detects whether there are vital signs around. When vital signs are detected around, the detective robot stops moving and takes a video in the direction where the vital signs exist, and marks the taken video as the detective video; The rescue determination module is used to determine whether to upload the detective video and the target terminal for video upload according to the video frames. Specifically: Convert the detective video into video frames, use the video frames as the input data of the physical sign detection model, obtain the output data of the physical sign detection model, mark the training label of the output data of the physical sign detection model as the physical sign detection value, sort all the physical sign detection values in the order of the corresponding video frames, calculate the difference between the adjacent subsequent physical sign detection value and the previous one after sorting to obtain the physical sign change value. When the physical sign change value > 0, mark this physical sign change value as the physical sign false change value, obtain the total physical sign false value Fy, set the physical sign detection threshold. When the physical sign detection value ≥ the physical sign detection threshold, mark this physical sign detection value as the detection false value, obtain the weakness detection value Sc, obtain the detective warning value Lw of this detective video, set the high detective warning value as Mb, and set the low detective warning value as Zj. When the detective warning value Lw ≥ the high detective warning value Mb, this detective robot immediately sends the detective video to the terminal of the rescue expert group, marks this detective robot as the expert rescue robot, and marks the position where this detective robot takes the video as the expert rescue position. When the low detective warning value Zj ≤ the detective warning value Lw < the high detective warning value Mb, this detective robot immediately sends the detective video to the terminal of the ordinary rescue group. When the detective warning value Lw < the low detective warning value Zj, send the position where this detective robot takes the video to the terminal of the ordinary rescue group, and control the detective robot to continue moving to detect vital signs; The supplementary rescue module is used to control the fixed rescue robot to move to the expert rescue position to take a video. Specifically: Take the expert rescue position of the expert rescue robot as the center, draw a circle with a preset radius to obtain the supplementary rescue range, mark the detective robots that are moving and whose positions are within the supplementary rescue range as the robots to be supplemented, obtain the supplementary rescue value Jq of the robots to be supplemented, mark the robot to be supplemented with the largest supplementary rescue value as the fixed rescue robot, control the fixed rescue robot to move to the expert rescue position, and the expert rescue robot and the fixed rescue robot take videos of the expert rescue position from multiple angles.

2. The detective robot control system for emergencies according to claim 1, wherein The physical sign detection model is obtained through the following steps: Obtain a plurality of image frames, label the image frames as training images, assign training labels to the training images, divide the training images into a training set and a validation set according to a set ratio, construct a neural network model, perform iterative training on the neural network model through the training set and the validation set. When the number of iterative training times is greater than the iterative number threshold, it is determined that the neural network model is trained. Label the trained neural network model as the physical sign detection model. The larger the training label value of the data output by the physical sign detection model, the weaker the vital signs in the image frame.

3. The detective robot control system for emergencies according to claim 2, characterized in that, The total virtual value Fy of the physical signs is obtained through the following steps: Mark the virtual change value of the physical signs as Ri, set the virtual change value coefficient of the physical signs as Pq, and use the formula to obtain the total virtual change value Gt of the physical signs. Here, Lm is the total number of virtual change values of the physical signs. The total virtual value Fy of the physical signs is obtained by using the formula Fy = Gt × a1 + Lm × a2, where a1 is the total virtual change value coefficient of the physical signs and a2 is the virtual change quantity coefficient of the physical signs.

4. The detective robot control system for emergencies according to claim 3, characterized in that, The weakness detection value Sc is obtained through the following steps: Calculate the difference between the detected weakness value and the physical sign detection threshold to obtain the detected weakness difference. Sum up all the detected weakness differences and take the average to obtain the average detected weakness difference, which is marked as Dz. Obtain the total number of times the physical sign detection value is marked as the detected weakness value, which is marked as Sg. Sort all the detected weakness values in the order of the corresponding video frames. Calculate the time difference between the times corresponding to two adjacent video frames after sorting to obtain the detected weakness value interval. Sum up all the detected weakness value intervals and take the average to obtain the average weakness value interval, which is marked as Pe. Use the formula to obtain the weakness detection value Sc, where b1 is the average detected weakness difference coefficient, b2 is the detected weakness value number coefficient, and b3 is the average weakness value interval.

5. The detective robot control system for emergencies according to claim 4, characterized in that The detective warning value Lw of the detective video is obtained through the following steps: Use the formula Lw = Fy × c1 + Sc × c2 to obtain the detective warning value Lw of the detective video, where c1 is the total physical sign weakness coefficient and c2 is the weakness detection value coefficient.

6. The detective robot control system for emergencies according to claim 5, characterized in that, The supplementary rescue value Jq of the robot to be supplemented is obtained through the following steps: Obtain the position of the robot to be supplemented, calculate the difference between the position of the robot to be supplemented and the expert rescue position to obtain the supplementary rescue distance, which is marked as Lk, obtain the remaining power of the robot to be supplemented, which is marked as Ns, and use the formula to obtain the supplementary rescue value Jq of the robot to be supplemented, where d1 is the supplementary rescue distance coefficient and d2 is the remaining power coefficient.

7. The detective robot control system for emergencies according to claim 6, characterized in that, When the physical sign change value ≤ 0, no processing is performed.

8. The detective robot control system for emergencies according to claim 7, characterized in that, When the physical sign detection value < the physical sign detection threshold, no processing is performed.

Images