CN113780072A - Fall detection method, system and computer-readable storage medium - Google Patents
Fall detection method, system and computer-readable storage medium Download PDFInfo
- Publication number
- CN113780072A CN113780072A CN202110882638.2A CN202110882638A CN113780072A CN 113780072 A CN113780072 A CN 113780072A CN 202110882638 A CN202110882638 A CN 202110882638A CN 113780072 A CN113780072 A CN 113780072A
- Authority
- CN
- China
- Prior art keywords
- mobile device
- obtaining
- human body
- environment image
- detection method
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 66
- 230000033001 locomotion Effects 0.000 claims abstract description 32
- 238000012545 processing Methods 0.000 claims abstract description 16
- 238000000034 method Methods 0.000 claims description 37
- 230000006870 function Effects 0.000 claims description 18
- 230000015654 memory Effects 0.000 claims description 18
- 238000011156 evaluation Methods 0.000 claims description 14
- 230000008859 change Effects 0.000 claims description 11
- 230000009471 action Effects 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 8
- 230000001629 suppression Effects 0.000 claims description 6
- 230000007613 environmental effect Effects 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 4
- 230000008569 process Effects 0.000 description 11
- 230000001133 acceleration Effects 0.000 description 8
- 230000004888 barrier function Effects 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000009467 reduction Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000135 prohibitive effect Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000007723 transport mechanism Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/60—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices
- G16H40/67—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for remote operation
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H80/00—ICT specially adapted for facilitating communication between medical practitioners or patients, e.g. for collaborative diagnosis, therapy or health monitoring
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
- Pathology (AREA)
- Alarm Systems (AREA)
Abstract
The invention discloses a tumble detection method, a tumble detection system and a computer-readable storage medium, wherein the tumble detection system comprises a mobile device and a control terminal, the mobile device is provided with a processing module, a driving module and an acquisition module, the processing module is respectively and electrically connected with the driving module and the acquisition module, and the tumble detection method comprises the following steps: acquiring the current motion state of the mobile device; acquiring an environment image from an acquisition module; planning a moving path of the mobile device according to the motion state and the environment image to obtain a tour path; driving the mobile device to move through the driving module according to the patrol path; obtaining an indication message for indicating that the environment image contains the human body information in the falling state according to the environment image; and sending the indication message to the control terminal. The mobile device moves on the tour route, continuously collects the environment images and detects the environment images, so that the condition in the area can be detected only by arranging the camera on the mobile device, and the detection cost is reduced.
Description
Technical Field
The invention relates to the technical field of computer vision recognition, in particular to a tumble detection method, a tumble detection system and a computer-readable storage medium.
Background
As society develops, many young people choose to compete in large cities, which also results in an increase in the number of empty nesters. When the old people fall down, if the old people cannot get timely help, life danger can be caused. Therefore, how to quickly find the fallen old people in the situation without other people becomes a crucial problem. In the related art, a detection method based on camera vision exists, and a large number of cameras are arranged in public places to acquire images, so that whether old people fall down or not is monitored. This approach, while effective, is cost prohibitive.
Disclosure of Invention
The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.
The embodiments of the present invention mainly aim to provide a fall detection method, a system and a computer-readable storage medium, by which detection cost can be reduced while ensuring completion of detection.
In a first aspect, an embodiment of the present invention provides a fall detection method applied to a fall detection system, where the system includes a mobile device and a control terminal, the mobile device is provided with a processing module, a driving module and an acquisition module, the processing module is electrically connected to the driving module and the acquisition module, respectively, and the method includes:
acquiring the current motion state of the mobile device;
acquiring an environment image from an acquisition module;
planning a moving path of the mobile device according to the motion state and the environment image to obtain a tour path;
driving the mobile device to move through the driving module according to the patrol path;
obtaining an indication message according to the environment image, wherein the indication message is used for indicating that the environment image contains human body information in a falling state;
and sending the indication message to the control terminal.
The fall detection method provided by the embodiment of the invention has at least the following beneficial effects: the moving path of the mobile device is planned through the motion state of the mobile device and the environment image from the acquisition module, and in the process of moving according to the planned patrol path, the detection is carried out according to the environment image, and whether a human body is in a falling state in the acquired environment image is judged. According to the method, the mobile device moves on the set tour path, the environment image is continuously collected and detected, so that the condition in one area can be detected only by arranging the camera on the mobile device, and the detection cost is reduced on the premise of ensuring the detection.
According to some embodiments of the first aspect of the present invention, the planning the moving path of the mobile device according to the motion state and the environment image to obtain the patrol path includes obtaining an actual speed range of the mobile device according to the motion state and the environment image, where the actual speed range includes an actual linear speed range and an actual angular speed range; sampling the linear velocity range and the angular velocity range to obtain a plurality of samples; predicting the action track of the mobile device according to the samples to obtain a plurality of predicted tracks; and scoring the plurality of predicted tracks according to an evaluation function to obtain a tour route. The patrol route is planned according to the actual motion capacity of the mobile device, so that the condition that accidents occur because the speed of the mobile device cannot meet the requirement of the patrol route is avoided.
According to some embodiments of the first aspect of the present invention, the obtaining, from the environment image, an indication message indicating that the environment image contains human body information in a fall state includes: extracting the features of the environment image to obtain human body feature maps with a plurality of scales; obtaining a human body regression frame according to the human body feature maps with multiple scales; obtaining a human body aspect ratio according to the human body regression frame; and when the human body aspect ratio reaches a preset threshold value, obtaining the indication message. The human body aspect ratio is used for judging whether the human body falls or not, and most of falling actions can be effectively judged.
According to some embodiments of the first aspect of the present invention, after the obtaining a human regression box according to the human feature maps of multiple scales, the method further includes: obtaining an effective area ratio and a central change rate according to the human body regression frame; and determining that the effective area ratio and the central change rate both reach preset thresholds. There is still some action that uses the aspect ratio of human body to judge that there is a possibility of error, so that the double judgment of effective area ratio and center change rate is added to ensure the correct detection.
According to some embodiments of the first aspect of the present invention, deriving an actual speed range of the mobile device from the motion state and the environment image, the actual speed range comprising an actual linear speed range and an actual angular speed range, comprises: obtaining a theoretical speed range of the mobile device according to the motion state, wherein the theoretical speed range comprises a theoretical linear speed range and a theoretical angular speed range; obtaining obstacle information according to the environment image; and obtaining the actual speed range of the mobile device according to the theoretical speed and the obstacle information. Due to the influence of environmental parameters, there may be a deviation between the actual speed of the mobile device and the theoretical speed. Therefore, on the basis of the theoretical speed, the environmental barrier is added as the weight to obtain the actual speed of the mobile device, so that the path planning is more accurate.
According to some embodiments of the first aspect of the present invention, the evaluation function includes a plurality of weight parameters, the plurality of weight parameters includes a deflection angle, a safety factor, and a time coefficient, the deflection angle is used to evaluate an angle difference between a terminal direction of the predicted trajectory and a target point, the safety factor is used to evaluate a probability of collision between the mobile device and an obstacle in the predicted trajectory, and the time coefficient is used to evaluate a time consumption of the predicted trajectory that can implement safe obstacle avoidance. And adding a plurality of weight parameters during path planning so as to obtain an optimal tour path.
According to some embodiments of the first aspect of the present invention, after the obtaining a human regression box according to the human feature maps of multiple scales, the method further includes: processing the human regression box using a non-maxima suppression method.
According to some embodiments of the first aspect of the present invention, after the obtaining a human regression box according to the human feature maps of multiple scales, the method further includes: and processing the human body regression frame according to a loss function.
In a second aspect, embodiments of the present invention provide a fall detection system, which includes a memory, a processor, a program stored on the memory and operable on the processor, and a data bus for implementing connection communication between the processor and the memory, wherein the program, when executed by the processor, implements the fall detection method according to the first aspect.
In a third aspect, the present invention provides a computer-readable storage medium storing one or more programs, which are executable by one or more processors to implement the fall detection method of the first aspect described above.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
Fig. 1 is a flow chart of a fall detection method according to an embodiment of the present invention;
fig. 2 is another flow chart of a fall detection method according to an embodiment of the present invention;
fig. 3 is another flow chart of a fall detection method according to an embodiment of the present invention;
fig. 4 is another flow chart of a fall detection method according to an embodiment of the present invention;
fig. 5 is another flow chart of a fall detection method according to an embodiment of the present invention;
fig. 6 is a schematic diagram of a dynamic window sampling trajectory according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
It should be noted that although the functional blocks are divided in the system architecture diagram, and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the blocks in the apparatus or in the flowchart.
The embodiments of the present invention will be further explained with reference to the drawings.
As shown in fig. 1, fig. 1 is a flowchart of a fall detection method according to an embodiment of the present invention. It can be appreciated that the present invention proposes a fall detection method, which is applied to a fall detection system. The tumble detection system comprises a mobile device and a control terminal, wherein the mobile device is provided with a processing module, a driving module and a collection module, and the processing module is connected to the driving module and the collection module respectively. The fall detection method includes, but is not limited to, step S100, step S200, step S300, step S400, step S500, and step S600.
Step S100, acquiring a current motion state of the mobile device.
Step S200, an environment image from an acquisition module is acquired.
And step S300, planning the moving path of the mobile device according to the motion state and the environment image to obtain a tour path.
And step S400, driving the mobile device to move through the driving module according to the patrol path.
And S500, obtaining an indication message according to the environment image, wherein the indication message is used for indicating that the environment image contains the human body information in the falling state.
Step S600, the indication message is sent to the control terminal.
It is understood that, in the present embodiment, the steps S100 and S200 are first performed, the current motion state of the mobile device is acquired, and the environment image acquired by the acquisition module is acquired. Then, step S300 is executed to plan a moving path of the mobile device according to the motion state and the environment image to obtain a patrol path, and then step S400 is executed to drive the mobile device to move according to the patrol path through the driving module. In the process of moving the mobile device, the acquisition module continuously acquires the environment image, and executes step S500 to obtain the indication message according to the environment image. The indication message is used for indicating that the environment image contains the human body information in the falling state. After the indication message is obtained, step S600 is executed, and the indication message is sent to the control terminal, so that the manager can quickly obtain the position of the fallen human body and timely help the fallen human body. In this embodiment, a moving path of the mobile device is planned through a motion state of the mobile device and an environment image from the acquisition module, and in a process of moving according to the planned patrol path, detection is performed according to the environment image, and whether a human body is in a falling state in the acquired environment image is judged. According to the method, the mobile device moves on the set tour path, the environment image is continuously collected and detected, so that the condition in one area can be detected only by arranging the camera on the mobile device, and the detection cost is reduced on the premise of ensuring the detection.
It should be noted that, in the process of path planning, the present invention uses a planning method based on a dynamic window. The core Dynamic window of a Dynamic Window Algorithm (DWA) is a window, which is a limited range by calculating the maximum and minimum linear and angular velocities of a current robot from the current state of a mobile device and a motion model of the mobile device. The current state of the mobile device refers to the current speed and the course angle of the mobile device, the motion model of the mobile device refers to the maximum linear speed, the maximum angular speed, the acceleration and the rotation acceleration which can be reached by the mobile device, the position which can be reached under each speed and angular speed is calculated in the window range, each position is evaluated (the evaluation content comprises the distance according to an obstacle, the angle towards the end point and the like), the current optimal position is selected, and then the optimal position is continuously repeated to establish a new window, so that a so-called dynamic window is formed.
It is understood that, in step S600, the indication message is sent to the control terminal, where the control terminal may refer to an application terminal of the mobile phone. When the mobile device detects and identifies the target area and finds that the old people fall down, the time and the position information of the falling down of the old people can be rapidly sent to the mobile phone application end through the Internet, and medical safety mechanisms or family members can be informed at the first time, so that the medical safety mechanisms or the family members can be rapidly rescued.
As shown in fig. 2, fig. 2 is another flow chart of a fall detection method according to an embodiment of the present invention. It is understood that step S300 in the embodiment shown in fig. 1 includes, but is not limited to, step S310, step S320, step S330, and step S340.
Step S310, obtaining an actual speed range of the mobile device according to the motion state and the environment image, wherein the actual speed range comprises an actual linear speed range and an actual angular speed range.
Step S320, sampling the linear velocity range and the angular velocity range to obtain a plurality of samples.
In step S330, the action trajectory of the mobile device is predicted according to the plurality of samples, so as to obtain a plurality of predicted trajectories.
And step S340, scoring the plurality of predicted tracks according to the evaluation function to obtain the tour route.
It can be understood that the tour route is planned according to the actual motion capability of the mobile device, so that the condition that accidents occur because the speed of the mobile device cannot meet the requirement of the tour route is avoided.
It should be noted that the evaluation function includes a plurality of weight parameters, where the weight parameters include, but are not limited to, a deflection angle, a safety factor, and a time coefficient, the deflection angle is used to evaluate an angle difference between a terminal direction of the predicted trajectory and the target point, the safety factor is used to evaluate a probability that the mobile device collides with the obstacle in the predicted trajectory, and the time coefficient is used to evaluate a time consumption of the predicted trajectory that can implement safe obstacle avoidance.
It is understood that after the predicted trajectories of the mobile device are obtained, the evaluation function is used to score each predicted trajectory, and the trajectory with the highest score is selected as the comprehensive tour path and executed. First, the following formula is processed:
G(v,w)＝α·corner(v,w)+β·security(v,w)+γ·velocity(v,w)
wherein, the corner (v, w) is a deflection angle evaluation subfunction, the subfunction has the function of evaluating the angle difference between the end direction of the predicted track and the target point at the predicted track speed, the formula is 180-theta (the smaller theta is, the higher the score is, wherein theta is the included angle between the end point of the predicted track and the connecting line between the mobile device and the target point), and the connecting subfunction mainly has the function of promoting the azimuth angle of the mobile device to continuously face the target point in the motion process; security (v, w) is a safety coefficient evaluation subfunction which is used for eliminating a predicted track which is likely to collide with an obstacle, so that the safe obstacle avoidance of the mobile device is realized. In order to avoid the condition that the proportion of the evaluation function is too large, when the predicted track without the barrier is evaluated, the safety coefficient evaluation sub-function is set as a constant; velocity (v, w) is a speed evaluation subfunction which is used for selecting a path with the highest speed from predicted tracks capable of realizing safe obstacle avoidance so as to reach a target point as soon as possible.
It should be noted that the dynamic patrol route planning process based on the dynamic window algorithm specifically includes the following steps:
(1) acquiring the positions of the mobile device and the obstacle in the environment image according to the acquisition module and constructing a simulation map;
(2) initializing a dynamic window algorithm and setting maximum linear velocity, minimum linear velocity, maximum angular velocity, minimum angular velocity, linear acceleration, angular acceleration, linear velocity resolution, angular velocity resolution, time resolution, track prediction time, barrier radius and weights of all evaluation subfunctions;
(3) updating a speed range according to the mechanical characteristics of the mobile device and the environment of the obstacle, and determining a dynamic window consisting of all feasible speeds;
(4) generating forward simulation time t of each sampling point in dynamic window according to motion state of mobile devicesimThe motion track of the mobile device;
(5) and (4) executing the optimal speed, checking whether the target point of the path is reached, if not, returning to the step (4), if so, outputting the result, and taking the predicted track as the finally determined tour path.
As shown in fig. 3, fig. 3 is another flow chart of a fall detection method according to an embodiment of the present invention. It is understood that step S500 in the embodiment shown in fig. 1 includes, but is not limited to, step S510, step S520, step S530 and step S540.
And step S510, extracting the features of the environment image to obtain human body feature maps with multiple scales.
And step S520, obtaining a human body regression frame according to the human body feature maps with multiple scales.
Step S530, obtaining the human body aspect ratio according to the human body regression frame.
And step S540, when the aspect ratio of the human body reaches a preset threshold value, obtaining an indication message.
It can be understood that the aspect ratio is used to judge whether the human body falls or not, and most of the falling actions can be effectively judged.
It can be understood that, because the characteristics of the human body proportion and the shape are easy to capture and the implementation of the algorithm is simple, the invention adopts the human body height ratio characteristic to judge the environment image. The human body width-height ratio can effectively judge most falling actions.
After step S520 is performed, the human body regression frame is obtained from the human body feature maps of a plurality of scales, and then the human body regression frame is further processed by using the non-maximum suppression method. In the detection process, firstly, the collected environment image is subjected to feature extraction through a backbone network, and the human body detection algorithm uses ShuffleNet as the backbone network; and then, respectively detecting by using feature maps of different scales, wherein anchor frames of different sizes and length-width ratios are defined on the feature maps of different scales in advance by the network, and the sizes of the anchor frames from the shallow feature map to the deep feature map are gradually increased, namely, the shallow feature map is used for predicting a small target, and the deep feature map is used for predicting a large target, so that the problem of target scale change in detection is solved. Finally, the network processes the detection results of the feature maps of different scales using a Non-Maximum Suppression (NMS) algorithm.
In the process of using the non-maximum suppression method, it is necessary to select a regression frame with the highest local score from the human body regression frames and suppress a human body regression frame with a lower local score using the non-maximum suppression method.
It can be understood that after the step S520 is performed, after the human body regression frame is obtained according to the human body feature maps of multiple scales, the human body regression frame needs to be further processed according to the loss function, and the loss function is formed by using the weighted sum of the confidence coefficient loss and the positioning loss. In the process of processing the human body regression frame by using the loss function, the loss value of the human body regression frame with the highest local score and the loss value of the image label need to be calculated by using the loss function, and then the quality of model prediction is evaluated. The processing procedure is shown in the following formula:
wherein x is a matching result of the default frame and the different types of group Truth frames; c is the confidence of the prediction box; l is the position information of the prediction frame; g is the position information of the ground truth frame; n is the default frame number; α is a parameter that trades off confidence loss and location loss and is typically set to 1.
It should be noted that the confidence loss is calculated as shown in the following formula:
wherein the content of the first and second substances,
p is the number of prediction categories; pos denotes positive samples, Neg denotes negative samples, Lconf(x, c) is confidence loss, and multiple classes of Softmax loss are used.
The positioning loss is calculated as follows:
wherein L isloc(x, L, g) for localization loss, Smooth L1 loss was used, and the above formula was satisfied
As shown in fig. 4, fig. 4 is another flow chart of a fall detection method according to an embodiment of the present invention. It is understood that, after step S520 in the embodiment shown in fig. 3, there are also included, but not limited to, step S550 and step S560.
And step S550, obtaining the effective area ratio and the center change rate according to the human body regression frame.
And step S560, determining that the effective area ratio and the center change rate both reach preset thresholds.
It will be appreciated that there is still some activity that uses the human aspect ratio to determine that a fault may occur, and therefore a double determination of the effective area ratio and the rate of change of the center is added to ensure correct detection. For the actions which are possibly misjudged by the aspect ratio of the human body, such as stretching exercise and squatting, the actions are eliminated by adopting the effective area ratio and the central change rate. In the process of detecting the human body regression frame, firstly, judging the human body regression frame, judging whether an aspect ratio threshold value is met, namely the aspect ratio is larger than 1, if the aspect ratio threshold value is not met, processing the next frame of image, if the aspect ratio threshold value is met, judging effective area ratio and center change, excluding the conditions of some special actions and activities, if all conditions are met, indicating that a person falls down, and finally outputting information and giving a warning. And if not, continuously acquiring the next frame of image for judgment.
As shown in fig. 5, fig. 5 is another flow chart of a fall detection method according to an embodiment of the present invention. It is understood that step S310 in the embodiment shown in fig. 2 further includes, but is not limited to, step S311, step S312, and step S313.
Step S311, obtaining a theoretical velocity range of the mobile device according to the motion state, where the theoretical velocity range includes a theoretical linear velocity range and a theoretical angular velocity range.
In step S312, obstacle information is obtained from the environment image.
In step S313, the actual speed range of the mobile device is obtained from the theoretical speed and the obstacle information.
It should be noted that, when sampling the linear velocity range and the angular velocity range, the following formula needs to be processed:
Vi＝{v∈[vmin,vmax]∩w∈[wmin,wmax]}
wherein, ViFor the theoretical velocity range of the solution, vminAnd vmaxMinimum and maximum linear velocity, w, of the moving means, respectivelyminAnd wmaxRespectively, a minimum angular velocity and a maximum angular velocity of the mobile device.
It should be noted that, under the influence of the driving module of the mobile device, the moment provided by the acceleration and deceleration is limited, so that a dynamic window exists in the period of simulating the forward movement of the mobile robot. Therefore, at a speed within the window, the actual speed that the mobile device can achieve under the influence of its own mechanical properties can be obtained by the following formula:
wherein, VjIs the actual speed, v, that the mobile device can achieve under the influence of its own mechanical propertiescIs the current linear velocity of the mobile device,
the maximum linear velocity of the mobile robot is added (subtracted); w is acThe current acceleration of the mobile robot;
the maximum plus (minus) angular velocity of the mobile robot; Δ t is the time increment.
It will be appreciated that there may be a deviation between the actual speed of the mobile device and the theoretical speed due to the influence of environmental parameters. Therefore, on the basis of the theoretical speed, the environmental barrier is added as the weight to obtain the actual speed of the mobile device, so that the path planning is more accurate. In order to realize safe obstacle avoidance and avoid collision with an obstacle occupying a certain space, the range of the speed can be obtained under the condition of deceleration and maximum acceleration as shown in the following formula, and the range of a dynamic window is further narrowed:
wherein distance (v, w) is the minimum Euclidean distance of the obstacle on the predicted trajectory of the corresponding velocity,
in order to achieve the maximum linear reduction speed after reduction,
for maximum angular speed reduction after reductionAnd (4) degree.
Respectively obtaining the theoretical speed range V of the mobile device according to the formulaiActual speed V that the mobile device can achieve under the influence of its own mechanical propertiesjAnd the range V of speeds that the mobile device can obtain under deceleration maximum acceleration conditionskThen, defining the dynamic window by the following formula, wherein V is the range of the dynamic window:
V＝Vi∩Vj∩Vk
as shown in fig. 6, fig. 6 is a schematic diagram of a dynamic window sampling trajectory according to an embodiment of the present invention. It will be appreciated that the motion trajectory is primarily based on the sampling point of each linear and angular velocity of the mobile device, and the forward simulation time tsimAnd (4) generating. According to the invention, the mobile device vision is combined with the lightweight human body detection algorithm and the falling detection algorithm for inspection, so that the falling detection speed of the old people is increased, the falling event can be found more quickly, and the efficiency of rescuing the empty-nest old people is improved. The dynamic path planning based on the dynamic window algorithm helps the mobile device to quickly reach a target detection area, and when the old people accidentally fall down, the old people can be effectively rescued in a short time, so that accidents are avoided.
In addition, another embodiment of the present invention also provides a fall detection system, including: a memory, a processor, and a computer program stored on the memory and executable on the processor.
The processor and memory may be connected by a data bus or other means.
The memory, which is a non-transitory computer readable storage medium, may be used to store non-transitory software programs as well as non-transitory computer executable programs. Further, the memory may include high speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory optionally includes memory located remotely from the processor, and these remote memories may be connected to the processor through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The non-transitory software programs and instructions required to implement the fall detection method of the above-described embodiment are stored in a memory, and when executed by a processor, perform the fall detection method of the above-described embodiment, for example, performing the above-described method steps S100 to S600 in fig. 1, method steps S310 to S340 in fig. 2, method steps S510 to S540 in fig. 3, method steps S550 to S560 in fig. 4, and method steps S311 to S313 in fig. 5.
The above-described embodiments of the apparatus are merely illustrative, wherein the units illustrated as separate components may or may not be physically separate, i.e. may be located in one place, or may also be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment.
Furthermore, an embodiment of the present invention also provides a computer-readable storage medium storing computer-executable instructions, which are executed by a processor or controller, for example, by a processor in the above-mentioned embodiment of the fall detection system, and can make the processor perform the fall detection method in the above-mentioned embodiment, for example, perform the above-mentioned method steps S100 to S600 in fig. 1, method steps S310 to S340 in fig. 2, method steps S510 to S540 in fig. 3, method steps S550 to S560 in fig. 4, and method steps S311 to S313 in fig. 5.
One of ordinary skill in the art will appreciate that all or some of the steps of the methods, systems, functional modules/units in the devices disclosed above may be implemented as software, firmware, hardware, and suitable combinations thereof.
In a hardware implementation, the division between functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed by several physical components in cooperation. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those of ordinary skill in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, Digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as known to those skilled in the art.
The preferred embodiments of the present invention have been described above with reference to the accompanying drawings, and are not to be construed as limiting the scope of the invention. Any modifications, equivalents and improvements which may occur to those skilled in the art without departing from the scope and spirit of the present invention are intended to be within the scope of the claims.
Claims (10)
1. The utility model provides a fall detection method, its characterized in that is applied to fall detection system, the system includes mobile device and control terminal, last processing module, the drive module and the collection module of installing of mobile device, processing module respectively with the drive module with collection module electric connection, the method includes:
acquiring the current motion state of the mobile device;
acquiring an environment image from the acquisition module;
planning a moving path of the mobile device according to the motion state and the environment image to obtain a tour path;
driving the mobile device to move through the driving module according to the patrol path;
obtaining an indication message according to the environment image, wherein the indication message is used for indicating that the environment image contains human body information in a falling state;
and sending the indication message to the control terminal.
2. The fall detection method according to claim 1, wherein said planning a movement path of said mobile device according to said motion state and said environmental image, obtaining a tour path, comprises:
obtaining an actual speed range of the mobile device according to the motion state and the environment image, wherein the actual speed range comprises an actual linear speed range and an actual angular speed range;
sampling the linear velocity range and the angular velocity range to obtain a plurality of samples;
predicting the action track of the mobile device according to the samples to obtain a plurality of predicted tracks;
and scoring the plurality of predicted tracks according to an evaluation function to obtain a tour route.
3. The fall detection method according to claim 2, wherein obtaining an indication message according to the environment image, the indication message being used to indicate that the environment image contains the human body information in the fall state, comprises:
extracting the features of the environment image to obtain human body feature maps with a plurality of scales;
obtaining a human body regression frame according to the human body feature maps with multiple scales;
obtaining a human body aspect ratio according to the human body regression frame;
and when the human body aspect ratio reaches a preset threshold value, obtaining the indication message.
4. The fall detection method according to claim 3, further comprising, after said obtaining a human regression box from said human feature maps of a plurality of scales:
obtaining an effective area ratio and a central change rate according to the human body regression frame;
and determining that the effective area ratio and the center change rate both reach preset thresholds.
5. The fall detection method according to claim 2, characterized in that said deriving from said movement state and said environment image an actual speed range of said mobile device, said actual speed range comprising an actual linear speed range and an actual angular speed range, comprises:
obtaining a theoretical speed range of the mobile device according to the motion state, wherein the theoretical speed range comprises a theoretical linear speed range and a theoretical angular speed range;
obtaining obstacle information according to the environment image;
and obtaining the actual speed range of the mobile device according to the theoretical speed and the obstacle information.
6. The fall detection method according to claim 2, wherein the evaluation function comprises a plurality of weight parameters, and the plurality of weight parameters comprise a deflection angle for evaluating an angle difference between a terminal direction of the predicted trajectory and a target point, a safety factor for evaluating a probability of collision of the mobile device with an obstacle in the predicted trajectory, and a time coefficient for evaluating a time consumption of the predicted trajectory for enabling safe obstacle avoidance.
7. The fall detection method according to claim 4, further comprising, after said obtaining a human regression box from said human feature maps of a plurality of scales:
processing the human regression box using a non-maxima suppression method.
8. The fall detection method according to claim 7, further comprising, after said obtaining a human regression box from said human feature maps of a plurality of scales:
and processing the human body regression frame according to a loss function.
9. A fall detection system, characterized in that it comprises a memory, a processor, a program stored on said memory and executable on said processor, and a data bus for implementing a connection communication between said processor and said memory, said program, when executed by said processor, implementing the steps of the fall detection method according to any one of claims 1 to 8.
10. A computer-readable storage medium, characterized in that it stores a computer-executable program for causing a computer to execute the fall detection method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110882638.2A CN113780072B (en) | 2021-08-02 | 2021-08-02 | Fall detection method, system and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110882638.2A CN113780072B (en) | 2021-08-02 | 2021-08-02 | Fall detection method, system and computer readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113780072A true CN113780072A (en) | 2021-12-10 |
| CN113780072B CN113780072B (en) | 2023-11-28 |
Family
ID=78836565
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110882638.2A Active CN113780072B (en) | 2021-08-02 | 2021-08-02 | Fall detection method, system and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113780072B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108837430A (en) * | 2018-07-05 | 2018-11-20 | 郑州大学 | A kind of multifunctional human balance evaluation and test training system |
| CN110032932A (en) * | 2019-03-07 | 2019-07-19 | 哈尔滨理工大学 | A kind of human posture recognition method based on video processing and decision tree given threshold |
| CN112330915A (en) * | 2020-10-29 | 2021-02-05 | 五邑大学 | Unmanned aerial vehicle forest fire prevention early warning method and system, electronic equipment and storage medium |
| CN112631294A (en) * | 2020-12-16 | 2021-04-09 | 上海应用技术大学 | Intelligent path planning method for mobile robot |
| CN113050646A (en) * | 2021-03-22 | 2021-06-29 | 西安工业大学 | Dynamic environment path planning method for indoor mobile robot |
-
2021
- 2021-08-02 CN CN202110882638.2A patent/CN113780072B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108837430A (en) * | 2018-07-05 | 2018-11-20 | 郑州大学 | A kind of multifunctional human balance evaluation and test training system |
| CN110032932A (en) * | 2019-03-07 | 2019-07-19 | 哈尔滨理工大学 | A kind of human posture recognition method based on video processing and decision tree given threshold |
| CN112330915A (en) * | 2020-10-29 | 2021-02-05 | 五邑大学 | Unmanned aerial vehicle forest fire prevention early warning method and system, electronic equipment and storage medium |
| CN112631294A (en) * | 2020-12-16 | 2021-04-09 | 上海应用技术大学 | Intelligent path planning method for mobile robot |
| CN113050646A (en) * | 2021-03-22 | 2021-06-29 | 西安工业大学 | Dynamic environment path planning method for indoor mobile robot |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113780072B (en) | 2023-11-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112418069B (en) | High-altitude parabolic detection method and device, computer equipment and storage medium | |
| CN115603466B (en) | Ship shore power system based on artificial intelligence visual identification | |
| CN112733690B (en) | High-altitude parabolic detection method and device and electronic equipment | |
| CN111611853A (en) | Sensing information fusion method and device and storage medium | |
| CN109803127A (en) | Urban safety building site monitoring system and method based on big data and technology of Internet of things | |
| CN108230617A (en) | Monitoring method for early warning and unmanned plane based on unmanned plane | |
| CN107031661A (en) | A kind of lane change method for early warning and system based on blind area camera input | |
| CN115909816A (en) | Buoy collision early warning and recording system | |
| CN108181905A (en) | A kind of obstacle avoidance method and system of pilotless automobile | |
| CN114155470A (en) | River channel area intrusion detection method, system and storage medium | |
| CN113780072B (en) | Fall detection method, system and computer readable storage medium | |
| CN113591574A (en) | Power transmission line inspection method and device based on laser radar | |
| CN117416349A (en) | Automatic driving risk pre-judging system and method based on improved YOLOV7-Tiny and SS-LSTM in V2X environment | |
| CN112489125A (en) | Automatic detection method and device for storage yard pedestrians | |
| CN114754744B (en) | Reservoir water level dynamic monitoring method based on computer image recognition | |
| CN112406700B (en) | Blind area early warning system based on upper and lower binocular vision analysis range finding | |
| CN113076797B (en) | Charging station electric vehicle fire alarm method and system based on intelligent video identification | |
| CN115273401A (en) | Method and system for automatically sensing falling of person | |
| CN113221800A (en) | Monitoring and judging method and system for target to be detected | |
| CN114782883A (en) | Abnormal behavior detection method, device and equipment based on group intelligence | |
| KR20220067271A (en) | Image acquisition apparatus and image acquisition method | |
| CN111339898A (en) | Behavior detection method and apparatus, computer readable storage medium, computer device | |
| CN116647032B (en) | Real-time protection system and method for power transmission line of target construction vehicle | |
| CN114078331B (en) | Overspeed detection method, overspeed detection device, visual sensor and storage medium | |
| CN108256509A (en) | A kind of fall detection method based on mobile phone camera |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |