CN111513731A

CN111513731A - Flexible intelligent detection device based on sufficient state monitoring attention

Info

Publication number: CN111513731A
Application number: CN202010347050.2A
Authority: CN
Inventors: 吴幸; 何倩; 沙正飞; 陈杰涛; 龙彦汐; 周文彬; 田希悦
Original assignee: East China Normal University
Current assignee: East China Normal University
Priority date: 2020-04-28
Filing date: 2020-04-28
Publication date: 2020-08-11
Anticipated expiration: 2040-04-28
Also published as: CN111513731B

Abstract

The invention discloses a flexible intelligent detection device based on attention monitoring of foot states, which comprises a flexible array type pressure sensor, an analog signal multiplexer, an analog-to-digital conversion circuit, a Bluetooth module, an artificial intelligent learning system and an insole base, wherein the flexible array type pressure sensor is connected with the analog signal multiplexer; the invention adopts a flexible array type pressure sensor to collect plantar pressure signals of dynamic or static sitting postures of a human body, the online pressure signals are input into an analog-to-digital conversion circuit through an analog signal multiplexer and are transmitted to a computer of an artificial intelligent learning system through a Bluetooth module, the online pressure signals are compared with standard pressure signals in a classifier by the computer, collected data are analyzed through a clustering algorithm, human behavior and attention quality are associated, real-time results are displayed in a display screen through signal identification and characteristic processing, and thus intelligent detection of human attention is realized. The invention has the advantages of high detection accuracy, small volume, light weight, convenient use and no limitation of regional environment.

Description

Flexible intelligent detection device based on sufficient state monitoring attention

Technical Field

The invention relates to the technical field of intelligent wearable devices, in particular to a flexible intelligent detection device based on attention monitoring of foot states.

Background

Attention refers to the ability of a person's mental activities to point at and focus on something. Attention is one of five basic factors of intelligence, namely the preparation states of memory, observation, imagination and thinking, so that attention is called a portal of soul. Attention is not a static psychological characteristic, but has the characteristics of high dynamic and fluctuating, and various human mental activities, including posture, emotion, volition and the like, need attention to participate so as to effectively occur. In the present period of research, it has been proved that attention influences the working efficiency of individuals or the performance of the academic industry, and the importance of studying the attention quality of individuals can be seen, and especially, the attention quality performance or the measurement of the students in an off-line class is worth exploring.

As technology develops, in addition to traditional attention observation: in addition to homework tests, empirical observations, etc., many people have been working on measuring attention qualities from instruments. The measuring points are divided into three types according to the measuring entry points, and the instrumental measuring methods respectively comprise the following steps: electroencephalography (EEG), visual, behavioral. The application number is 201620006749.1, the patent name is a student classroom participation detection system, and discloses a student classroom participation detection system, which comprises a student intelligent terminal, a camera for attendance, an attendance computer, a camera for attention detection and a video processing computer; the intelligent terminal is used for scanning the two-dimensional code generated by the check-in computer to complete check-in and automatically log in a classroom interaction system, and monitoring and analyzing the participation degree of students in a classroom by a method of measuring through a visual instrument.

According to the principle of an instrument for measuring brain waves, according to the principle that alpha wave bands in individual brain waves mainly move at 8-12 Hz in a quiet state and an awake state, the frequency of frontal lobe brain waves of an individual with attention defects is reduced, the power of theta slow waves is increased, the ratio of theta/beta power is increased, then the brain waves are measured and recorded through sensor equipment, and the brain waves are converted into a popular and understandable data form through an algorithm, so that the attention is measured. Such instruments are bulky, expensive and are only used in small scale experimental environments.

In the vision measurement method, a camera is used for collecting information of a learner in a learning process, and a mental state of an individual is analyzed by analyzing facial expressions through a computer algorithm (such as a face recognition technology), so that the attention condition of the individual is analyzed. Eye tracking techniques, which analyze individual attentiveness from eye annotation location (attention allocation), annotation time (task difficulty and amount of attention resources needed), and scan path, and Kinect are important examples of visual measurements; the Kinect is used as a 3D somatosensory camera, and recognizes human body images, voice and the like to analyze psychological states so as to obtain attention information. The instrument has higher time cost in data acquisition and processing and has high requirements on data storage.

In the behavior measuring method, basic material equipment is mainly used, and the behavior measuring method has the characteristics of low cost and high popularity. In the current products for measuring attention through behaviors, attention information is mainly obtained according to the principle that attention concentration is different in different sitting postures, and the sitting postures are analyzed by the methods of sitting posture inclination and sitting posture pressure measurement. The sitting posture inclination is to acquire the inclination condition of a body by using a gyroscope, the pressure measurement is to acquire the pressure distribution condition of a cushion by using a pressure sensor, the two mechanisms are to acquire sitting posture information such as sitting posture information of sitting, forward leaning, backward leaning, side forward (backward) leaning and the like, and different sitting postures represent different attention qualities. However, this type of method has a great problem, and the test is affected by the special sitting postures of the legs of the Erlang and the pan legs, and the use scenario has great limitations.

In 1993 Teasdale et al, it was reported that when the plantar pressure of a human is closer to the arch of the foot, the posture is more stable, and less attention is required to be paid to the mechanism for regulating the posture stability, that is, more attention can be focused on other things. Armstrong and Edgley discovered in 1984 that single support costs more attention from animal and human gait, and we can therefore compare the attention of the legs of the dirached with that of other sitting postures. Studies in 2015 at the university of four colleges in the united states at the time of analysis of ADHD for children indicate that this involuntary repetitive leg shaking action helps to improve concentration and concentration when focusing on doing a task. Thus, attention is paid to the fact that different sitting postures are accompanied by different foot pressure distribution conditions. In addition, the frequency of limb movement is also strongly related to attention, for example, in the test of hyperactivity disorder (ADHD) and attention, it can be confirmed that a symptom of a patient with hyperactivity disorder is that the patient often feels uneasy, plays hands or feet, or continuously twists. We can be through observing the frequency of foot pressure analysis position of sitting condition and limbs change to acquire the attention condition, the wearable shoe-pad of flexible intelligence can reach this purpose, and have stronger convenience and portability, overcome the influence of the leg of the two langs, dish leg simultaneously.

Most of the prior art are expensive, have high requirements on scenes and equipment, do not consider the influence of the change frequency of limbs on attention, are not accurate and effective enough and are not convenient enough to detect the attention, and the sufficient pressure is closely related to the sitting posture, and the sitting posture and the sufficient pressure condition influence the attention, so that the intelligent flexible device is designed, and the attention is analyzed by detecting the sufficient pressure condition.

Disclosure of Invention

The invention aims to provide a flexible intelligent detection device for monitoring attention based on a foot state, which aims at overcoming the defects of the prior art, and is characterized in that a sole pressure signal of a dynamic or static sitting posture of a human body is acquired through a flexible array type pressure sensor, the online pressure signal is input into an analog-to-digital conversion circuit through an analog signal multiplexer and is transmitted to a computer of an artificial intelligent learning system through a Bluetooth module, the online pressure signal is compared with a standard pressure signal in a classifier through the computer, the acquired data is analyzed through a clustering algorithm, the behavior of the human body is associated with the attention quality, and a real-time result is displayed in a display screen through signal identification and characteristic processing, so that the intelligent detection of the attention of the human body is realized.

The specific technical scheme for realizing the purpose of the invention is as follows:

a flexible intelligent detection device based on attention monitoring of foot states is characterized by comprising a flexible array type pressure sensor, an analog signal multiplexer, an analog-to-digital conversion circuit, a Bluetooth module, an artificial intelligent learning system and an insole base;

the artificial intelligence learning system consists of a computer, a classifier and a display screen, and the computer is connected with the classifier and the display screen through data lines;

the insole base is divided into an upper cushion layer and a lower device layer;

the flexible array type pressure sensor is arranged on a cushion layer of the insole base, and the analog signal multiplexer, the analog-to-digital conversion circuit and the Bluetooth module are arranged on a device layer of the insole base;

the computer, the classifier and the display screen of the artificial intelligence learning system are sequentially and independently arranged;

the flexible array type pressure sensor, the analog signal multiplexer, the analog-to-digital conversion circuit and the Bluetooth module are sequentially connected through leads, and the Bluetooth module is in wireless communication connection with a computer of the artificial intelligent learning system.

The invention collects plantar pressure signals of dynamic or static sitting postures of a human body through the flexible array type pressure sensor, the online pressure signals are input into the analog-to-digital conversion circuit through the analog signal multiplexer and are transmitted to the computer of the artificial intelligent learning system through the Bluetooth module, the online pressure signals are compared with standard pressure signals in the classifier through the computer, the collected data are analyzed through a clustering algorithm, the behaviors of the human body are related to the attention quality, and real-time results are displayed in a display screen through signal identification and characteristic processing, so that the intelligent detection of the attention of the human body is realized.

Drawings

FIG. 1 is a schematic structural view of the present invention;

FIG. 2 is a schematic view of the assembled structure of the insole base;

FIG. 3 is a circuit schematic of an analog signal multiplexer;

FIG. 4 is a schematic diagram of a structure of an artificial intelligence learning system;

FIG. 5 is a schematic diagram of ten static sitting postures of a human body;

FIG. 6 is a schematic diagram illustrating an embodiment of the present invention.

Detailed Description

Referring to fig. 1, the invention comprises a flexible array type pressure sensor 1, an analog signal multiplexer 2, an analog-to-digital conversion circuit 3, a bluetooth module 4, an artificial intelligence learning system 5 and an insole base 6;

the artificial intelligence learning system 5 is composed of a computer 51, a classifier 52 and a display 53, and the computer 51 is connected with the classifier 52 and the display 53 through data lines;

the insole base 6 is divided into a left insole base layer 61 and a right insole base layer 62;

the flexible array type pressure sensor 1 is arranged on a cushion layer 61 of the insole base 6, and the analog signal multiplexer 2, the analog-to-digital conversion circuit 3 and the Bluetooth module 4 are arranged on a device layer 62 of the insole base 6;

the computer 51, the classifier 52 and the display 53 of the artificial intelligence learning system 5 are sequentially and independently arranged;

the flexible array type pressure sensor 1, the analog signal multiplexer 2, the analog-to-digital conversion circuit 3 and the Bluetooth module 4 are sequentially connected through leads, and the Bluetooth module 4 is in wireless communication connection with a computer 51 of the artificial intelligent learning system 5.

The invention works as follows:

referring to fig. 1 and 6, to detect the sitting posture of the person, two soles are placed on the flexible array pressure sensors 1 of the insole base 6, under the pressure action of each point of the soles, a plurality of groups of pressure signals of the two soles obtained by the flexible array pressure sensors 1 are input to the analog-to-digital conversion circuit 3 through the analog signal multiplexer 2, and are transmitted to the computer 51 of the artificial intelligent learning system 5 through the bluetooth module 4, the computer 51 compares the online pressure signals with the standard pressure signals in the classifier 52, and real-time results are displayed in the display screen 53 through signal identification and feature processing, so that the intelligent detection of the attention of the person is realized.

Referring to fig. 1, 5 and 6, it is found from the research on the relationship between human behavior and attention that different sitting postures of the human body keep close relationship with the attention, and to realize the real-time detection of the attention of students, the present invention first collects the static and dynamic pressure signals of different sitting postures of the human body as standard pressure signals to be input into the classifier 52, and divides the standard pressure signals of different static sitting postures of the human body into ten states of sitting upright, left inclined, right inclined, leg tilting with the palm upward, flat leaning foot, flat leaning tiptoe landing, flat leaning foot, flat leaning heel landing, and crossed heel tilting, and the dynamic standard pressure signals are two states of leg shaking and posture fluctuation.

Referring to fig. 1, 5 and 6, because the sole of the person to be detected is attached to the flexible array pressure sensor 1, the flexible array pressure sensor 1 collects the online pressure signals of each part of one or two soles in real time, the online pressure signals are input to the analog-to-digital conversion circuit 3 through the analog signal multiplexer 2 and are wirelessly transmitted to the computer 51 of the artificial intelligent learning system 5 through the bluetooth module 4, the computer 51 identifies and compares the online pressure signals with the standard pressure signals in the classifier 52, and the real-time results are displayed in the display screen 53 through signal identification and feature processing, so that the intelligent detection of the attention of the human body is completed.

Compared with the prior art, the invention has the advantages of high detection accuracy, small volume, light weight, convenient use and no limitation of regional environment, overcomes the defect that the existing scene type, image type and wearable type detection have strict requirements on equipment and environment, can improve the learning attention of students by sitting posture correction on the premise of not interfering the normal learning and life style of the person to be detected, and provides reliable and effective analysis data for the correlation research of human behavior and attention.

Examples

Referring to fig. 1 and 2, the flexible array type pressure sensor 1 of the present invention has 64 rows and 64 columns of sensing arrays selected, and is composed of 64 × 64 RX-D2027 flexible pressure sensors, which are arranged on the cushion layer 61 of the insole base 6 to collect pressure values of each contact of the foot as online pressure signals; the measuring range and the size of the flexible array type pressure sensor 1 can be set according to the weight and the foot length of a user; each sensor array is distributed at each position of the foot, so that different pressure values of each point of the foot are obtained, the plurality of sensor arrays work simultaneously, the stm32 supplies power to the whole system in a circulating manner, the analog signal multiplexer 2 is used for providing working voltage to the sensor arrays in turn, the gating positions are switched at a high speed, and the purpose of continuously acquiring the pressure values of each part of the sole is achieved; the analog signal multiplexer 2, the analog-to-digital conversion circuit 3 and the Bluetooth module 4 are arranged on the device layer of the insole base 6, and the online pressure signal is wirelessly transmitted to the computer 51 of the artificial intelligent learning system 5 through the Bluetooth module 4.

Referring to fig. 1 and 3, the analog signal multiplexer 2 of the present invention uses two TP0164 analog signal 64 multiplexers to control 64 rows and 64 columns, respectively, the analog signal multiplexer functions as an analog switch, and controls the output of one of the sensors by changing the gating position at a high speed, and since the output of the corresponding sensor is converted at a high speed, the analog signal multiplexer approximates to that of each sensor to continuously output data;

the analog signal multiplexer 2 realizes the output measurement of the sensor array, the analog switch has low power consumption, high speed and low on-resistance, and the on-resistance is kept relatively stable in the whole input signal range.

Referring to fig. 1, 2 and 5, the pressure value of each pressure sensor is output to the analog-to-digital conversion circuit 3 in serial after being gated, and converted into a digital signal convenient for the computer 51, each analog-to-digital converter of the analog-to-digital conversion circuit 3 shares up to 16 external channels, the conversion time can reach 1 μ s, the analog signals of ten kinds of static and two kinds of dynamic pressure data of the human subject 6 can be converted into digital signals, the analog-to-digital conversion circuit 3 adopts ADC multi-channel scan conversion of 2 12-bit analog-to-digital converters embedded in stm32, performs bluetooth communication through a USART serial port of stm32 in an embedded system in a DMA transmission mode, and is connected to the bluetooth module 4 to transmit the packaged data to the computer 51.

Referring to fig. 1 and 4, the artificial intelligence learning system 5 of the present invention is composed of a computer 51, a classifier 52 and a display 53, wherein in the artificial intelligence learning system 5, a clustering algorithm program is set in the computer 51; various standard pressure signals are provided in the classifier 52, including a dynamic or static classification standard pressure signal, a dynamic sitting posture standard pressure signal, and a static sitting posture standard pressure signal.

Referring to fig. 1, 4 and 6, the artificial intelligence learning system 5 of the present invention comprises the following three steps:

firstly, classifying dynamic and static sitting postures: the computer 51 receives a group of data transmitted by Bluetooth every 0.05s, each group of data contains online pressure signals of corresponding coordinates of each sensor array, the computer 51 receives the online pressure signals transmitted by the Bluetooth module 4, the computer 51 slidingly stores two online pressure signals with the duration of 100s to form two online pressure signal data packets, the online pressure signal data packets are compared with dynamic or static classification standard pressure signals in the classifier 52, the computer 51 firstly analyzes and confirms that the signals belong to dynamic sitting posture signals or static sitting posture signals from the online pressure signal data packets, if the signals are confirmed to be static sitting posture signals, the confirmed static sitting posture signals are sent to the classifier 52, and the classifier 52 matches the corresponding static sitting posture standard pressure signals;

secondly, classifying specific sitting postures: the computer 51 identifies and compares the first online pressure signal data packet with the standard static sitting posture pressure signal in the classifier 52, analyzes the static sitting posture type with the highest matching degree, and makes a first sitting posture classification determination for the sole online pressure signal data packet; the computer 51 identifies and compares the second online pressure signal data packet with the standard static sitting posture pressure signal in the classifier 52, analyzes the static sitting posture type with the highest matching degree, and makes a second sitting posture classification determination on the sole online pressure signal data packet; if the two classification results are inconsistent, the online pressure signal data packet needs to be collected again, and if the two classification results are consistent, the next step is carried out;

thirdly, analyzing attention and outputting a detection result through the sitting posture: the computer 51 compares and corrects the two online pressure signal data packets to obtain an attention quality data packet, analyzes and calculates the attention quality data packet through a clustering algorithm, namely analyzes that an attention quality parameter is output, the full score is 100, the higher the score is, the more concentrated the corresponding attention condition is, outputs the analyzed sitting posture and the analyzed attention quality parameter to the display screen 53, and displays a real-time result in the display screen 53, thereby realizing the intelligent detection from the state of foot pressure to the human attention.

Claims

1. A flexible intelligent detection device based on attention monitoring of foot states is characterized by comprising a flexible array type pressure sensor (1), an analog signal multiplexer (2), an analog-to-digital conversion circuit (3), a Bluetooth module (4), an artificial intelligent learning system (5) and an insole base (6);

the artificial intelligence learning system (5) is composed of a computer (51), a classifier (52) and a display screen (53), and the computer (51) is connected with the classifier (52) and the display screen (53) through data lines;

the insole base (6) is arranged at the left and right, and the insole base (6) is divided into an upper cushion layer (61) and a lower device layer (62);

the flexible array type pressure sensor (1) is arranged on a cushion layer (61) of the insole base (6), and the analog signal multiplexer (2), the analog-to-digital conversion circuit (3) and the Bluetooth module (4) are arranged on a device layer (62) of the insole base (6);

the computer (51), the classifier (52) and the display screen (53) of the artificial intelligence learning system (5) are sequentially and independently arranged;

the flexible array type pressure sensor (1), the analog signal multiplexer (2), the analog-to-digital conversion circuit (3) and the Bluetooth module (4) are sequentially connected through leads, and the Bluetooth module (4) is in wireless communication connection with a computer (51) of the artificial intelligent learning system (5).