TWI422824B - Human motion identification and locating method - Google Patents

Description

Human motion feature recognition and localization method

The invention relates to a human body motion feature recognition and positioning method, in particular to a human body motion feature recognition and positioning method using an inertial measurement device.

Most of the current mobile positioning technologies use signals from a global positioning system (GPS) or a wireless local area network (wireless local area network) to provide positioning data. However, when the GPS is blocked by buildings or forests, The positioning signal cannot be received; the positioning of the wireless LAN will also seriously affect the accuracy due to the complexity of the space geometry (such as a large number of tourists).

Inertial positioning detects the acceleration of the moving object to calculate its displacement, so it is not affected by the external environment. The higher the accuracy of the accelerometer's detection of the acceleration of a moving object, the more expensive it is.

Since the accelerometer measures the acceleration to produce an error, the displacement error obtained by quadratic integration of the acceleration in time will be more amplified with time.

In addition, inertial positioning that is not affected by the external environment is mostly used in aircraft, automobiles, and other machinery, because the acceleration and deceleration of its motion is obvious and smooth. For inert actors with motion patterns (eg animals, humans, etc.), traditional inertial positioning cannot be applied.

An embodiment of the present invention provides a human body motion feature recognition method for identifying a motion state of a user, including: receiving a first child inertial measurement device that is worn over a waist of the user. An acceleration and a first angular velocity; receiving a second angular velocity transmitted by a second inertial measurement device mounted on a lower leg or an ankle of the user; and a change according to the first acceleration and the first acceleration The amount is determined by the user's state of motion.

Another embodiment of the present invention provides a human body motion feature recognition and positioning method, including: measuring an acceleration in a vertical direction by a first inertial measurement device worn on a waist or waist of a user and obtaining a first a sensing signal and measuring an angular velocity in a horizontal direction and obtaining a second sensing signal; performing a heel detection procedure according to the first sensing signal and the second sensing signal; according to the first sensing signal Performing a behavior recognition program with the second sensing signal to generate a behavior recognition result; storing a plurality of data between the heel strikes; receiving a second worn on the user's calf or an ankle a third sensing signal transmitted by the inertia measuring device; calculating a vertical angular velocity according to the third sensing signal; calculating a moving distance of the user from the vertical angular velocity according to the data and generating a positioning information.

The above and other technical contents, features and advantages of the present invention will be apparent from the following detailed description of the preferred embodiments. The directional terms mentioned in the following embodiments, such as up, down, left, right, front or back, etc., are only directions referring to the additional drawings. Therefore, the directional terminology used is for the purpose of illustration and not limitation.

Fig. 1 is a schematic view showing the wearing of the inertial measurement device required for the human body motion feature recognition and positioning method according to the present invention. In the present invention, the human motion feature recognition and positioning method is completed by using two initial measurement units (IMUs), wherein a first inertial measurement device must be worn on the upper body (waist or waist) of the user. Above position), such as wearing positions 11 and 12. A second inertial measurement device is worn on the user's lower leg or ankle, such as the wearing position 13. The first inertial measuring device is configured to measure a vertical acceleration of the user and a first angular velocity in a horizontal direction for determining a motion state of the user. The second inertial measurement device is configured to measure a second angular velocity of the user in a vertical direction to determine the moving distance of the user as a reference for positioning. The state of motion includes going up the stairs, going down the stairs, standing up, sitting down, etc. The inertial measurement device can be covered by a housing.

Fig. 2 is a flow chart showing a method for recognizing human motion characteristics according to the present invention. In step S21, an electronic device receives a first acceleration and a first angular velocity transmitted by a first inertial measurement device that is worn over the waist or waist of the user. In step S22, the electronic device receives a second angular velocity transmitted by a second inertial measurement device that is worn on the user's lower leg or ankle. The first acceleration is a vertical acceleration of the user, the first angular velocity is the horizontal angular velocity of the user, and the second angular velocity is an angular velocity in the vertical direction. In step S23, signal conversion is performed on the received first acceleration, the first angular velocity and the second angular velocity, and the manner of signal conversion may be wavelet conversion or analog digital signal conversion. Next, in step S24, a motion state of the user is determined based on the first acceleration, the first angular velocity, and the second angular velocity.

Figure 3 is a schematic illustration of an embodiment of an inertial measurement device in accordance with the present invention. The inertia measurement device 31 is an inertial measurement device that is worn over the waist or waist of the user, and includes an acceleration sensor 32, an angular velocity sensor 33, a control unit 34, and a transmission unit 35. The acceleration sensor 32 is configured to sense a vertical acceleration of the user and generate a first sensing signal. The angular velocity sensor 33 is configured to sense an angular velocity of the user in a horizontal direction and generate a second sensing signal. The control unit 34 receives the first sensing signal and the second sensing signal, generates a vertical acceleration value and a horizontal angular velocity value, and transmits the signal to a receiving device through the transmitting unit 35. The sending unit 35 communicates with the receiving device by using a wireless network interface, a Bluetooth transmission interface, an infrared interface, a wireless RF interface or other non-contact connection.

Figure 4 is a schematic illustration of another embodiment of an inertial measurement device in accordance with the present invention. The inertia measurement device 41 is an inertial measurement device worn on the user's lower leg or ankle, and includes an angular velocity sensor 42, a control unit 43, and a transmission unit 44. The angular velocity sensor 42 is configured to measure an angular velocity of a vertical direction of the user and generate a third sensing signal. The control unit 43 receives and converts the third sensing signal to a vertical angular velocity value and transmits it to a receiving device through the transmitting unit 44. The sending unit 44 communicates with the receiving device by using a wireless network interface, a Bluetooth transmission interface, an infrared interface, a wireless RF interface or other non-contact connection.

Fig. 5 is a flow chart showing an embodiment of a human body motion feature recognition method according to the present invention. In step S51, a vertical acceleration is first measured by a first inertial measuring device worn on the waist or waist of the user, and a first sensing signal is obtained. Next, in step S52, the first sensing signal is subjected to signal conversion, such as wavelet signal conversion, to obtain a vertical acceleration value. In step S53, it is determined whether the vertical acceleration value is greater than a first threshold value. If so, it can be determined that the user's motion state is a stair state at this time, and if no, the process proceeds to step S55. In step S54, an antero-posterior acceleration is measured and a second sensing signal is obtained. Next, in step S55, the second sensing signal is subjected to signal conversion, such as wavelet signal conversion, to obtain an intermediate acceleration value. In step S56, it is determined whether the intermediate acceleration value is greater than a second threshold, and if so, it can be determined that the user's motion state is a walking state. If not, the process proceeds to step S57, where a motion mode comparison program is executed. In step S57, a plurality of comparison rules are used to determine the current state of motion of the user. In the embodiment, the current comparison state of the user is determined by the preset comparison rule to be the state of going up the stairs and standing up. Or sit down, but the invention is not limited thereto. The standing state refers to a state in which the user originally sat and then stood up. The sitting state refers to the state in which the user originally stood and then sat down.

Fig. 6 is a flow chart showing an embodiment of a method for recognizing and locating a human body motion feature according to the present invention. In step S61, a vertical acceleration is measured by a first inertial measuring device worn on the waist or waist of the user, and a first sensing signal is obtained and a horizontal angular velocity is measured and a second sensing is obtained. signal. Next, in step S62, noise filtering is performed on the first sensing signal and the second sensing signal. In step S63, the first sensing signal filtered by the noise and the second sensing signal are subjected to signal conversion, such as wavelet conversion or analog-to-digital conversion. Signal processing is then performed on the converted signal in step S64 to generate a plurality of processed values. In step S65, a heel strike detection program is executed to detect the user's walking state as a reference for calculating the user's walking distance.

In step S66, the converted signal is received and a behavior recognition program is executed to identify the current motion state of the user. In step S67, a recognition result of the behavior recognition program is received. In step S68, based on the identification result and the heel strike detection result, the distance traveled between the heel strikes and the user is calculated.

For a clearer explanation, please refer to Figure 7. Figure 7 is a flow diagram of an embodiment of heel landing detection in accordance with the present invention. In step S71, a sensing signal corresponding to an acceleration transmitted from an inertial measuring device worn on the user's lower leg or ankle is received. In step S72, the sensing signal is sampled, and the sampling frequency may be 50 Hz. Next, in step S73, noise filtering is performed on the sampled signal. In step S74, a detected value is calculated based on the filtered noise. In step S75, it is determined whether the detected value is greater than a critical value. In this embodiment, the threshold is 0.5 and the range of detected values is between 0 and 1. If the detected value is not greater than 0.5, the process returns to step S71. If the detected value is greater than 0.5, a heel strike signal is generated, and the data between the heel strike signals is stored twice as a reference for subsequent calculation of the user's moving distance.

Figure 8 is a flow diagram of an embodiment of a user positioning method in accordance with the present invention. In step S81, the user's data between the heel and the ground is received twice. Next, in step S82, the moving distance of the user is calculated. Next, in step S83, a direction of the user is calculated based on an angular velocity value transmitted from an inertial measurement device worn above the waist or waist of the user and an initial direction of the user. In step S84, a map correction procedure is performed. The map correction program determines whether the user's direction of travel changes in a large direction, such as a turn, based on the direction and distance of the user. If it is found, the user will clear the previous record data and re-receive the new data through the program. As a result, the accuracy of the positioning of the user can be improved. Next, in step S85, a coordinate information is converted according to an initial position, a moving distance and a direction of the user, and in step S86, the position of the user is displayed on the map according to the coordinate information.

Figure 9A is a schematic diagram of an embodiment of calculating the distance traveled by a user. The angles φ1 and φ2 can be calculated by an angular acceleration transmitted by an inertial measuring device worn on the user's lower leg or ankle. Φ1 is the angle between a vertical line and the front leg in the vertical ground direction. Φ2 is the angle between a vertical line and the rear leg in the vertical ground direction. In this embodiment, the user's step size is obtained by using a mathematical model of human motion, and the mathematical model of the human motion is as follows:

d _{mod el} = f ( L , P , Φ 1, Φ 2, θ )= N ( l sin Φ 1+ L sin Φ 2+ P sin θ )

The parameters are as follows:

Φ1: the angle between a vertical line and the front leg in the vertical ground direction

Φ2: the angle between a vertical line and the rear leg in the vertical ground direction

θ: angle of pelvic rotation

N: The number of heels hitting the ground

L, P: leg length and pelvic width

Figure 9B is a schematic diagram of a user going upstairs. In order to calculate the moving distance of the user upstairs, another mathematical model is used to calculate the moving distance of the user. The mathematical model is as follows:

d=(L-0.18)tan( Φ 1+ Φ 2)

Figure 9C is a schematic diagram of a user going downstairs. In order to calculate the moving distance of the user downstairs, another mathematical model is used to calculate the moving distance of the user. The mathematical model is as follows:

d=1.7 P sin( Φ 1+ Φ 2)

It should be noted that when the user goes up and down the stairs, the angle between the rear foot and the vertical line in the vertical ground direction should be larger than the angle between the front foot and the vertical line in the vertical ground direction. That is, φ2>φ1. Therefore, such a feature can be utilized to design a decision mechanism to avoid calculating the information that received the error. Please refer to Figure 9D. Figure 9D is a schematic illustration of an embodiment of a step size computing device in accordance with the present invention. The step size calculation unit 91 includes an angle calculation device 92, a determination unit 93, and a step size calculation module 94. The angle calculating device 92 receives an angular acceleration transmitted by an inertial measuring device worn on the lower leg or the ankle of the user, and calculates an angle φ1 between a vertical line in the vertical ground direction and the front leg of the user, and a vertical line in the vertical ground direction. The angle φ2 with the user's hind legs. The judging unit 93 first compares the magnitudes of the angles φ1 and φ2. If φ1 is larger than φ2, it means that the calculation result of the angle calculating means 92 is incorrect, and the calculation result is not transmitted to the step size calculation module 94. If φ1 is smaller than φ2, it means that the calculation result of the angle calculating means 92 is correct, and the calculation result is transmitted to the step size calculation module 94 to calculate the step size of the user.

Figure 10 is a flow chart of an embodiment of a user direction calculation method. In step S101, the user is given an initial direction. Next, in step S102, an angular velocity value transmitted by an inertial measuring device worn on the waist or waist of the user is received, and a steering angle of the user is calculated. Next, in step S103, a current direction of the user is calculated based on the initial direction and the steering angle.

The above is only the preferred embodiment of the present invention, and the scope of the invention is not limited thereto, that is, the simple equivalent changes and modifications made by the scope of the invention and the description of the invention are All remain within the scope of the invention patent. In addition, any of the objects or advantages or features of the present invention are not required to be achieved by any embodiment or application of the invention. In addition, the abstract sections and headings are only used to assist in the search of patent documents and are not intended to limit the scope of the invention.

11, 12, 13‧‧‧ wearing position

31‧‧‧Inertial measuring device

32‧‧‧Acceleration sensor

33‧‧‧Angle speed sensor

34‧‧‧Control unit

35‧‧‧Send unit

41‧‧‧Inertial measuring device

42‧‧‧Angle speed sensor

43‧‧‧Control unit

44‧‧‧Send unit

91‧‧‧Step calculation unit

92‧‧‧Angle computing device

93‧‧‧judging unit

94‧‧‧Step calculation module

Fig. 1 is a schematic view showing the wearing of the inertial measurement device required for the human body motion feature recognition and positioning method according to the present invention.

Fig. 2 is a flow chart showing a method for recognizing human motion characteristics according to the present invention.

Figure 3 is a schematic illustration of an embodiment of an inertial measurement device in accordance with the present invention.

Figure 4 is a schematic illustration of another embodiment of an inertial measurement device in accordance with the present invention.

Fig. 5 is a flow chart showing an embodiment of a human body motion feature recognition method according to the present invention.

Fig. 6 is a flow chart showing an embodiment of a method for recognizing and locating a human body motion feature according to the present invention.

Figure 7 is a flow diagram of an embodiment of heel landing detection in accordance with the present invention.

Figure 8 is a flow diagram of an embodiment of a user positioning method in accordance with the present invention.

Figure 9A is a schematic diagram of an embodiment of calculating the distance traveled by a user.

Figure 9B is a schematic diagram of a user going upstairs.

Figure 9C is a schematic diagram of a user going downstairs.

Figure 9D is a schematic illustration of an embodiment of a step size computing device in accordance with the present invention.

Figure 10 is a flow chart of an embodiment of a user direction calculation method

Claims (4)

A human motion feature recognition and positioning method includes: measuring a vertical acceleration by a first inertial measurement device worn on a waist or waist of a user and obtaining a first sensing signal and measuring one And a second sensing signal is obtained at a corner speed in the horizontal direction; performing a heel detection procedure according to the first sensing signal and the second sensing signal; performing according to the first sensing signal and the second sensing signal a behavior recognition program to generate a behavior identification result; storing a plurality of data between the heel strikes; receiving a second inertial measurement device worn on a lower leg or an ankle of the user a three sensing signal; calculating a vertical angular velocity according to the third sensing signal; Calculating a moving distance of the user according to the vertical angular velocity and generating a positioning information; setting an initial direction of the user; generating a vertical acceleration according to the first sensing signal; and according to the second sensing signal Generating a horizontal angular velocity; calculating a steering angle of the user based on the horizontal angular velocity; and generating a current direction of the user based on the initial direction and the steering angle. The method for identifying and locating a human motion feature according to claim 1, wherein the positioning information is determined according to the initial direction and the current direction of the user. Identification of human motion characteristics as described in item 1 of the patent application scope The positioning method includes: performing a sampling process on the first sensing signal and the second sensing signal to generate a first sampling signal and a second sampling signal; according to the first sampling signal The second sampled signal generates a plurality of detected values; and when the detected value is greater than a threshold, a heel strike signal is generated. The method for identifying and locating a human motion feature according to claim 3, wherein the threshold is 0.5.