CN210836549U

CN210836549U - Intelligent wearable sensing system for protecting distraction walking safety

Info

Publication number: CN210836549U
Application number: CN201922068716.6U
Authority: CN
Inventors: 肖世军
Original assignee: Individual
Current assignee: Individual
Priority date: 2019-11-27
Filing date: 2019-11-27
Publication date: 2020-06-23
Anticipated expiration: 2029-11-27

Abstract

The utility model discloses a sensing system is dressed to intelligence of protection distraction walking safety, include at least wearing sensing equipment and machine learning module with intelligent Mobile terminal wireless connection, dress sensing equipment and supply the user to dress and gather user surrounding environment characteristic information, intelligent Mobile terminal carries out radio communication with wearing sensing equipment, acquire environment characteristic data, intelligent Mobile terminal has the machine learning model that has been and study and judge environment characteristic data, remind the user to notice danger in the surrounding environment, machine learning module builds machine learning model by the server, and train machine learning model by data processing module, the server regularly updates machine learning model, the server carries out wireless network communication with intelligent Mobile terminal, intelligent Mobile terminal downloads and updates machine learning model. The utility model discloses the environment around the real-time perception of help user effectively solves pedestrian's distraction walking and leads to the attention to descend, take place incident scheduling problem easily.

Description

Intelligent wearable sensing system for protecting distraction walking safety

Technical Field

The utility model belongs to the technical field of artificial intelligence and specifically relates to a sensing system is dressed to intelligence of protection distraction walking safety.

Background

With the increasing popularity of smart phones, more and more people use the phones while walking, such as watching news, getting on social networks, playing games, navigating maps, and the like. When a person focuses on the screen of the mobile phone, his visual perception of the surroundings can be reduced to less than 10% of normal, while other perceptions such as hearing are also reduced considerably. As shown in fig. 1, when walking with distraction, not only the cars running around are dangerous, but also the cars may be stumbled by obstacles on steps or roads, step into a pit, hit a traffic sign or a pillar, hit another person walking with distraction, and the like. In areas and streets where urban vehicles are dense, the potential safety hazards for these distracted people are more severe. With the increase of smart phone users, the potential safety hazard of walking with distraction becomes an increasingly serious problem. Traffic accidents or safety accidents caused by the use of the smart phone for walking with distraction are increasing every year. The use of the smart phone is the decision of the pedestrian, and it is not practical to forbid the distracted walking through the traffic regulations, but an effective prevention and control means is still lacked in the prior art.

SUMMERY OF THE UTILITY MODEL

The applicant aims at the problems that the attention of the existing pedestrians is reduced and safety accidents are easy to happen due to the fact that the existing pedestrians walk with the smart phone in a distracted mode, and the like, and provides an intelligent wearable sensing system which is reasonable in structure and can protect the safety of the distracted walking, the danger contained in the surrounding environment can be reminded when the users walk with the distracted mode, the possibility of accidents is reduced, continuous learning and optimization can be achieved through a machine learning model, and the intelligent walking robot has the characteristic of machine intelligence.

The utility model discloses the technical scheme who adopts as follows:

the utility model provides a protect intelligent wearing sensing system of distraction walking safety, include at least wearing sensing equipment and machine learning module with intelligent mobile terminal wireless connection, wear sensing equipment and supply the user to dress and gather user's surrounding environment characteristic information, intelligent mobile terminal carries out radio communication with wearing sensing equipment, acquire environment characteristic data, intelligent mobile terminal studies and judges environment characteristic data with existing machine learning model, remind the user to pay attention to danger in the surrounding environment, machine learning module constructs machine learning model by the server, and train machine learning model by data processing module, the server regularly updates machine learning model, the server carries out wireless network communication with intelligent mobile terminal, intelligent mobile terminal downloads and updates machine learning model.

As a further improvement of the above technical solution:

the wearable sensing device includes a clothing worn or worn on a person's body and one or more sensors fixedly or movably connected to the clothing.

One or more sensors on the wearing article are made into an independent sensing piece, the shape of the sensing piece is a cubic block, and the back of the sensing piece is provided with a magnetic clamping piece.

The top of the sensing piece is provided with a hanging ring, or the magnetic clamping piece is in a concave curved surface.

One or more sensors on the wearing object are made into an independent sensing piece, the left side and the right side of the sensing piece are respectively connected with a section of elastic transparent adhesive tape, and the transparent adhesive tape on the sensing piece is wound on the intelligent mobile terminal and is bound.

The sensor is one or more of a visual sensor, an ultrasonic sensor, a sound sensor, an infrared laser radar and a radio frequency radar.

The wearing form of wearing sensing equipment is intelligent cap, intelligent clothes, intelligent bracelet, intelligent glasses, intelligent earphone or the combination of above form.

And a small solar cell panel is arranged on the intelligent cap and used for supplying power to the arranged sensor.

The intelligent mobile terminal is a smart phone, a palm computer or a smart watch.

The intelligent mobile terminal is internally provided with an application program, the intelligent mobile terminal is connected with the wearable sensing equipment and the machine learning module through the application program and carries out data transmission, the intelligent mobile terminal calculates by using a machine learning model stored by the intelligent mobile terminal after receiving the environmental characteristic data, and the intelligent mobile terminal transmits the original data to the machine learning module through a wireless network to carry out data storage and model optimization.

The utility model has the advantages as follows:

the utility model discloses remind with the help of the danger that contains in the surrounding environment when machine learning model walks with the help of the distraction of user, reduce the possibility that the occurence of failure, machine learning model carries out continuous study and optimization moreover, makes the utility model becomes intelligent system, uses artificial intelligence to user's distraction walking on, the environment around the real-time perception of help user effectively solves the pedestrian and uses smart mobile phone distraction walking to lead to attention to descend, take place incident scheduling problem easily. The intelligent mobile terminal timely reminds the user of possible danger through the installed application program, can detect objects or situations which bring safety problems to pedestrians through wearing sensing equipment, such as moving vehicles, steps, pits and barriers, and reminds the user through various modes, so that safety of the user during walking with distraction is guaranteed. The utility model discloses utilize data processing module to train machine learning model, continuous optimization machine learning model to improve the degree of accuracy that its discernment was judged, utilize artificial intelligence's continuous study to realize the continuous perfect of system, compare with the prior art that does not have learning ability, have the progress that is showing. The utility model provides a machine learning module can be served to numerous users anytime and anywhere based on the service pattern of cloud computing, dresses wearing that sensing equipment can be convenient moreover, and the sensor is with low costs, connects and operates based on the smart mobile phone, has fine practical value. The utility model discloses dispose the sensor on wearing articles such as cap, clothes and constitute and dress sensing equipment, do not change the original effect of wearing the article, with the help of wearing the article and personally surveying, it is more convenient.

Drawings

Fig. 1 is a schematic diagram of walking with the smart phone being distracted.

Fig. 2 is a schematic view of the structure of the present invention.

Fig. 3 is a schematic view of the working process of the present invention.

Fig. 4 is a schematic view of a wearable sensing device.

Fig. 5 is a schematic diagram of a smart cap.

Fig. 6 is a schematic view of a sensing member.

FIG. 7 is a schematic view of another sensing element.

Fig. 8 is a schematic view of yet another sensing element.

Fig. 9 is a schematic diagram of the present invention when a reminder is sent.

FIG. 10 is a schematic diagram of a neural network model.

In the figure: 1. wearing a sensing device; 2. an intelligent mobile terminal; 3. a machine learning module; 4. a sensor; 5. a server; 6. a data processing module; 7. a storage module; 8. a solar panel; 9. an intelligent cap; 10. an intelligent garment; 11. a smart bracelet; 12. smart glasses; 13. an intelligent earphone; 14. a sensing element; 15. a magnetic clip; 16. transparent adhesive tape.

Detailed Description

The following describes embodiments of the present invention with reference to the drawings.

As shown in fig. 2 and fig. 3, the intelligent wearable sensing system for protecting the walking security comprises a wearable sensing device 1, an intelligent mobile terminal 2 and a machine learning module 3, wherein the wearable sensing device 1 is worn by a user and has a sensor 4 for collecting characteristic information of the surrounding environment of the user, the intelligent mobile terminal 2 wirelessly communicates with the wearable sensing device 1 to obtain the characteristic data of the environment, the intelligent mobile terminal 2 uses an existing machine learning model and studies and judges the characteristic data of the environment to remind the user of danger in the surrounding environment, the machine learning module 3 constructs the machine learning model by a server 5, and the data processing module 6 trains the machine learning model, the server 5 updates the machine learning model regularly, the server 5 performs wireless network communication with the intelligent mobile terminal 2, and the intelligent mobile terminal 2 downloads and updates the machine learning model. The machine learning module 3 further comprises a storage module 7 for storing the raw data and the machine learning model. The machine learning module 3 is a service mode based on cloud computing, and constructs a machine learning cloud, and realizes corresponding functions through cloud computing.

The wearable sensing device 1 comprises a wearing article worn or worn on the human body and one or more sensors 4 fixedly or movably connected with the wearing article, such as one or more of a visual sensor, an ultrasonic sensor, a sound sensor, an infrared laser radar and a radio frequency radar, wherein the collected environment characteristic data is image data, video data, audio data, positioning data, radar data, ultrasonic reflection data, model data or combination data of the above data. The wearable sensing device 1 and the intelligent mobile terminal 2 establish wireless communication through a wireless transmission technology, such as bluetooth, WiFi or other internet of things technologies. Wearing sensing equipment 1 initiative discernment user environment around to gather surrounding environment characteristic information, generate environment characteristic data, send environment characteristic data to intelligent Mobile terminal 2 through radio communication.

As shown in fig. 4, the wearable sensing device 1 can be worn on different parts of the user's body in a form including, but not limited to, a smart cap 9, a smart garment 10, a smart band 11, smart glasses 12, a smart headset 13, or a combination thereof. For example, the smart cap 9, the smart garment 10, the smart headset 13, or the smart glasses 12 are provided with visual and auditory sensors, and the smart bracelet 11 is provided with an auditory sensor. Wherein, dispose a plurality of visual sensors on the intelligence cap 9, a plurality of visual sensors distribute in the different position on the cap, monitor one or more directions of user, gather the environmental data in a visual angle or the space visual angle. The smart cap 9 may also be provided with a small solar panel 8 for powering the sensor 4 provided (see fig. 5). When the sensors 4 are provided on the article of clothing that is removably attached to the article of clothing, one or more of the sensors 4 on the article of clothing may be formed as a separate sensing element 14, which sensing element 14 may be attached to or removed from the article of clothing. When the sensing element 14 is removed, it can be movably fixed on the smart mobile terminal 2, for example, the sensing element 14 is shaped as a housing with a connector, can be attached to a smart phone, and the like. When the sensors 4 configured on the wearing article are fixedly connected with the wearing article, one or more sensors 4 on the wearing article may be sewn, adhered, attached or integrated with the wearing article, such as one or more visual sensors sewn on the smart cap 9.

As shown in fig. 6, the sensing element 14 is in the shape of a cubic block, and has an ultrasonic sensor, a video sensor (camera) and an acoustic sensor (microphone) integrated therein, and the three sensors collect the ambient characteristic information from the front of the sensing element 14. The sensing element 14 is provided with a magnetic clip 15 on the back side, and the sensing element 14 can be mounted on a garment or a tie to form the smart garment 10. A hanging ring is also provided on the top of sensing element 14, and sensing element 14 can be worn around the neck of the user when the hanging ring is provided with a hanging chain. As shown in fig. 7, the back of the sensing element 14 can also be designed as a concave curved surface, and correspondingly, the magnetic clip 15 is also designed as a curved shape, and the intelligent hat 9 is formed by mounting the magnetic clip 15 on the hat. As shown in fig. 8, the sensing element 14 is a cubic block, and has an ultrasonic sensor, a video sensor (camera) and an acoustic sensor (microphone) integrated therein, the left and right sides of the sensing element 14 are respectively connected with a section of elastic transparent adhesive tape 16, the transparent adhesive tape 16 on the sensing element 14 is wound on the smart mobile terminal 2 and bound, and the three sensors collect the surrounding environment information from the top of the sensing element 14.

The intelligent mobile terminal 2 is an electronic device such as a smart phone, a palm computer or a smart watch, which can be used by a user in a distracted way during walking, an application program is installed in the intelligent mobile terminal 2, and the intelligent mobile terminal is connected with the wearable sensing device 1 and the machine learning module 3 through the application program and carries out data transmission. After receiving the environmental characteristic data, the intelligent mobile terminal 2 transmits the environmental characteristic data to the machine learning module 3 for model optimization periodically through a wireless network, stores the machine learning model in the intelligent mobile terminal 2, and downloads the updated machine learning model in time. The application program of the intelligent mobile terminal 2 utilizes a machine learning model to study and judge the environmental characteristic data: if the environmental characteristic data accords with the set conditions for danger during walking in the machine learning model, the intelligent mobile terminal 2 gives out danger alarm to the user to remind the user of the danger, such as information of running automobiles, speed, direction, distance and the like near the user. The reminding method includes, but is not limited to, displaying images and characters on a screen of the intelligent mobile terminal 2, and sending voice prompt, vibration prompt and the like through the intelligent mobile terminal 2 (see fig. 9). If the environmental characteristic data does not accord with the set conditions for danger occurrence in walking in the machine learning model, the intelligent mobile terminal 2 does not send out danger alarm, and continues to conduct research and judgment of the next environmental characteristic data. The wearable sensing device 1 can scan once every certain time period, for example, every second, so that the intelligent mobile terminal 2 can timely remind about potential danger.

The machine learning module 3 constructs a machine learning model by the server 5, the server 5 receives the environment characteristic raw data and stores the environment characteristic raw data in the storage module 7 for the server 5 to call at any time, such as data marking, and then the raw data are processed and converted into a machine learning format and then sent to the data processing module 6. The data processing module 6 has a powerful computing system including, but not limited to, an image processor. The data processing module 6 trains the machine learning model to optimize the machine learning model, such as optimizing a neural network. The optimized model file is stored in the storage module 7 by the server 5 for downloading or next optimization. The intelligent mobile terminal 2 may automatically download updates on a regular basis or the system may prompt the user to download an update model.

The algorithms of the machine learning model are support vector machine, decision tree, expectation maximization algorithm, artificial neural network or ensemble learning, etc., and the artificial neural network is taken as an example for explanation. The artificial neural network simulates the human cranial nerves, and the calculation process is similar to the human cranial nerve perception process. As shown in fig. 10, the artificial neural network has a plurality of input nodes and one or more output nodes, and is formed by interconnecting basic node units, and the artificial neural network has an ordered layer structure including an input layer, an intermediate layer, and an output layer, where the number of intermediate layers may be one or more, and the larger the number of intermediate layers, the better the recognition effect of the network is, which is called deep learning. The description parameters of the neural network model are recorded and stored in a file form, and comprise the number of network layers, the number of nodes of each layer, an excitation function of each node, the interconnection weight of all nodes and the like. For the identification of different objects, one larger network model with multiple outputs may be used, or smaller network models with a single output may be used, i.e. each object to be identified has a corresponding model.

The working process of the artificial neural network is as follows: a group of characteristic variable numerical value signals are extracted from the acquired image or sound data and input into an input layer of a neural network, the signals are subjected to linear weighted superposition, each interconnected path has a weight parameter and is input into the next node, and the output of the node is obtained through nonlinear excitation function calculation. Through the operation of all layers of the neural network, each layer carries out weighted superposition and excitation function numerical value calculation, and finally a probability value is output, such as the probability of a car in an image. And judging the probability of existence of a certain object or phenomenon in the environmental characteristic data by using the neural network model, and triggering alarm reminding if the existence probability exceeds a set reference value. The user downloads the neural network model file from the machine learning module 3 through the intelligent mobile terminal 2, and after the obtained environmental characteristic data is calculated and judged through the neural network model, the forward propagation is called. Taking the example of the vision sensing, the vision sensor scans a series of image data obtained by focusing at different forward focuses, such as 5, 10, 15, and 20 meters, and predicts the automobile by using an output of the neural network, such as calculating the probability of the existence of the automobile to be 80%, 95%, 80%, and 60%, respectively. If the reference value for the presence of the vehicle is set to 90%, it can be judged that the vehicle is approximately at 10 meters. To increase the distance accuracy, one image may also be scanned every 1 meter to find the distance corresponding to the maximum probability. In addition, the distance and the moving speed of the objects to the user can be measured by combining with an ultrasonic sensor, and when the distance between the automobile and the user is not within a safe range, the intelligent mobile terminal 2 can send out an alarm prompt, for example, when the automobile moves about 10 meters ahead, the speed is about 30 kilometers per hour.

The data processing module 6 trains the machine learning model and continuously optimizes the neural network so as to improve the intelligent accuracy of the neural network. The optimization of the neural network needs a large amount of matrix operation, powerful image processor resources are used, a large amount of environment characteristic data are needed for optimization, the data are marked, the optimization algorithm is called backward propagation, and the model is continuously optimized to finally obtain the optimal solution of parameters, so that the prediction error given by the model is minimum. When the neural network is optimized by using the marked data regularly, the accuracy of the identification judgment of the neural network is continuously improved, namely, the guided machine learning is realized.

The utility model discloses remind with the help of the danger that contains in the surrounding environment when machine learning model walks with the help of the user distraction, machine learning model carries out continuous study and optimization moreover, makes the utility model discloses become intelligent system. The intelligent mobile terminal 2 can timely remind the user of possible danger through the installed application program, can detect objects or situations which bring safety problems to pedestrians through the wearable sensing device 1, such as moving vehicles, steps, pits and barriers, and remind the user through various modes, so that safety of the user during walking with distraction is guaranteed. The utility model provides a machine learning module 3 can serve numerous users anytime and anywhere based on the service pattern of cloud computing, dresses the wearing that sensing equipment 1 can be convenient moreover, and the sensor is with low costs, connects and operates based on the smart mobile phone, has fine practical value.

The above description is illustrative of the present invention and is not intended to limit the present invention, and the present invention may be modified in any manner without departing from the spirit of the present invention.

Claims

1. The utility model provides a protection distraction walking safety's intelligence dress sensing system which characterized in that: the intelligent mobile terminal comprises at least a wearable sensing device (1) and a machine learning module (3) which are in wireless connection with an intelligent mobile terminal (2), wherein the wearable sensing device (1) is worn by a user and collects characteristic information of the surrounding environment of the user, the intelligent mobile terminal (2) is in wireless communication with the wearable sensing device (1) to obtain environmental characteristic data, the intelligent mobile terminal (2) uses an existing machine learning model to study and judge the environmental characteristic data to remind the user of danger in the surrounding environment, the machine learning module (3) constructs the machine learning model by a server (5), the machine learning model is trained by the data processing module (6), the server (5) updates the machine learning model regularly, the server (5) performs wireless network communication with the intelligent mobile terminal (2), and the intelligent mobile terminal (2) downloads and updates the machine learning model.

2. The intelligent wearable sensing system for protecting distraction walking safety according to claim 1, wherein: the wearable sensing device (1) comprises a wearing article worn or worn on the body of a person and one or more sensors (4) fixedly or movably connected with the wearing article.

3. The intelligent wearable sensing system for protecting distraction walking safety according to claim 2, wherein: one or more sensors (4) on the clothing are made into a separate sensing piece (14), the sensing piece (14) is in the shape of a cubic block, and the back of the sensing piece (14) is provided with a magnetic clamping piece (15).

4. The intelligent wearable sensing system for protecting distraction walking safety according to claim 3, wherein: the top of the sensing piece (14) is provided with a hanging ring, or the magnetic clamping piece (15) is in a concave curved surface.

5. The intelligent wearable sensing system for protecting distraction walking safety according to claim 2, wherein: one or more sensors (4) on the wearing article are made into an independent sensing piece (14), the left side and the right side of the sensing piece (14) are respectively connected with a section of elastic transparent adhesive tape (16), and the transparent adhesive tape (16) on the sensing piece (14) is wound on the intelligent mobile terminal (2) and is bound.

6. The intelligent wearable sensing system for protecting distraction walking safety according to claim 2, wherein: the sensor (4) is one or more of a visual sensor, an ultrasonic sensor, a sound sensor, an infrared laser radar and a radio frequency radar.

7. The intelligent wearable sensing system for protecting distraction walking safety according to claim 1, wherein: the wearing form of wearing sensing equipment (1) is intelligent cap (9), intelligent clothes (10), intelligent bracelet (11), intelligent glasses (12), intelligent earphone (13) or the combination of above forms.

8. The intelligent wearable sensing system for protecting distraction walking safety of claim 7, wherein: a small solar cell panel (8) is arranged on the intelligent cap (9) and used for supplying power to the arranged sensor (4).

9. The intelligent wearable sensing system for protecting distraction walking safety according to claim 1, wherein: the intelligent mobile terminal (2) is an intelligent mobile phone, a palm computer or an intelligent watch.

10. The intelligent wearable sensing system for protecting distraction walking safety according to claim 1, wherein: the intelligent mobile terminal (2) is internally provided with an application program, the intelligent mobile terminal is connected with the wearable sensing equipment (1) and the machine learning module (3) through the application program and carries out data transmission, the intelligent mobile terminal (2) calculates by using a machine learning model stored by the intelligent mobile terminal after receiving the environmental characteristic data, and the intelligent mobile terminal (2) transmits the original data to the machine learning module (3) through a wireless network to carry out data storage and model optimization.