Precise and rapid detection device and method for the center of gravity of wheelchair occupant's body posture

Abstract

The present invention discloses a precise and rapid detection device for the center of gravity of a wheelchair occupant's body posture, comprising a plurality of thin film pressure sensors and a corresponding acquisition card system, wherein the thin film pressure sensors are arranged at the seat cushion and the backrest of the wheelchair, the acquisition card system comprises an MCU controller and an acquisition card circuit, the thin film pressure sensors are connected to the acquisition card system, and a voltage comparator circuit is arranged in the acquisition card circuit; the present invention adopts sensors with a relatively small area and a large number to measure the center of gravity of a human body in a sitting posture, when the acquisition card system is in a working state, the voltage comparator circuit sends the state of each node to the MCU, and the sensor acquisition card converts the analog quantity of the voltage into a digital quantity through the voltage comparator circuit, so that it can be quickly judged whether pressure is applied to the node according to a set threshold value; and a scanning algorithm for the outer contour of the pressure area is designed to avoid scanning a large number of invalid nodes, thereby improving the efficiency of sensor scanning detection.

Description

Technical Field

[0001] The present invention belongs to the technical field of wheelchairs, and in particular relates to a precise and rapid detection device and method for the body posture and center of gravity of a wheelchair occupant. Background Art

[0002] In today's society, people's living conditions are constantly improving, and the average life expectancy of our country's citizens is also constantly increasing. However, as they age, the physical functions of the elderly will gradually weaken, especially the lower limb strength will decrease significantly, which has caused great difficulties for the normal life of the elderly. On the other hand, lower limb disabilities caused by congenital defects or acquired accidents have also brought great troubles to the lives of some people. As a new type of rehabilitation and disability aid, the smart wheelchair is suitable for people who have lost their mobility or have difficulty in moving. It can enable them to regain the ability to move independently. In the process of using the wheelchair, in order to improve its safety and operability, the following difficulties need to be overcome:

[0003] (1) If the wheelchair is in motion and the occupant has an improper sitting posture, he or she may fall or tip over, resulting in injury; if he or she does not receive timely treatment, his or her life may be in danger.

[0004] (2) When the occupant needs to stand up or has not sat down firmly, the wheelchair needs to remain stable and in a braking state to protect the occupant. Since it is impossible to accurately and quickly obtain the occupant's sitting posture information, this function cannot be promoted in the current wheelchair. By analyzing the above difficulties, it can be found that the sitting posture information of the wheelchair occupant is of great significance to the safety of wheelchair use. If the sitting posture information of the occupant can be detected quickly and accurately, the state of the wheelchair can be adjusted in real time according to different sitting postures, thereby realizing auxiliary control of the wheelchair. The current methods for detecting human sitting posture include but are not limited to the following:

[0005] 1. The document "A human sitting posture detection system based on a flexible array pressure-sensitive cushion" (patent, announcement number: CN108549834A) introduces a pressure-sensitive cushion detection device, which is mainly composed of three parts: a flexible array pressure-sensitive cushion, an embedded acquisition circuit, and a host computer software, and is used to detect and identify various sitting postures of daily desk workers. The acquisition circuit of this method does not process the collected voltage signal, but sends the data frame by frame to the host computer for calculation. Due to the huge amount of data, it relies on the calculation and processing of the PC, and cannot achieve accurate and rapid judgment of the user's sitting posture under a smaller system scale.

[0006] 2. The document "Sitting Posture Detection System Based on a Combination of Multiple Sensors" (Science and Technology Innovation, July 2021) is mainly composed of a flexible pressure-sensitive sensor array, a pressure signal processing circuit and an infrared ranging sensor, which are used to measure relevant sitting posture data to identify bad sitting postures. The specific acquisition scheme is: when collecting data from the flexible pressure-sensitive array sensor, the voltage analog output obtained by the series voltage division of the resistor is directly connected to the 16-channel ADC of the microcontroller. This detection method occupies a large amount of IO resources of the microcontroller, has a slow detection speed and low efficiency, and is only suitable for experimental measurement of sitting posture data. It cannot quickly and accurately collect and judge the user's sitting posture.

[0007] In order to simplify the detection device and increase the detection speed, some literature also introduces the use of a small number of large pressure-touch films to detect the sitting posture of the human body. However, a small number of pressure sensors may have uneven force, resulting in a large deviation when calculating the center of gravity of the human body based on the center of the pressure-touch sensor, which is not precise enough. Summary of the invention

[0008] To solve the above problems, the present invention discloses a precise and rapid detection device and method for the center of gravity of a wheelchair occupant's body posture, which adopts a relatively large number of sensors with a relatively small area to measure the center of gravity of a sitting human body, and converts the analog value of the voltage into a digital value through a voltage comparator circuit of a sensor acquisition card, so that it can quickly determine whether pressure is applied to a node according to a set threshold value; and a scanning algorithm for the outer contour of the pressure area is designed to avoid scanning a large number of invalid nodes, thereby improving the efficiency of sensor scanning detection.

[0009] To achieve the above object, the technical solution of the present invention is as follows:

[0010] A precise and rapid detection device for the center of gravity of a wheelchair occupant's body posture comprises a plurality of thin film pressure sensors and a corresponding acquisition card system. The thin film pressure sensors are arranged on the seat cushion and the backrest of the wheelchair. The acquisition card system comprises an MCU controller and an acquisition card circuit. The thin film pressure sensors are connected to the MCU controller via a control line. A voltage comparator circuit is arranged in the acquisition card circuit.

[0011] Furthermore, the number of acquisition nodes of the thin film pressure sensor is 24×24. When the nodes on the thin film pressure sensor are subjected to pressure, the resistance values ​​at both ends thereof will drop from several megohms to several thousand ohms or even lower. Based on this characteristic, the pressure can be determined by measuring the resistance value.

[0012] Furthermore, the control line 1 and the control line 2 of the MCU controller control the electronic switch 1 and the electronic switch 2 respectively, forming a gating circuit for connecting the resistors connected to different rows and columns to the same-proportional operation circuit; the same-phase proportional operation circuit is mainly composed of an operational amplifier and a feedback resistor, and when the resistance value in the access circuit changes, its output voltage value will also change accordingly. In addition to the above basic circuit for collecting the resistance value of the sensor, the present invention also adds a voltage comparator circuit for quickly judging the resistance value of the sensor node. The voltage comparator circuit is composed of a voltage comparator chip and a potentiometer, and the potentiometer is used to adjust the set voltage threshold; the input voltage from the operational amplifier will be compared with the threshold voltage, and when the threshold voltage is reached, the voltage comparator circuit outputs a high level, otherwise it outputs a low level, and the level signal is connected to the IO port of the MCU controller to collect the status information of each node.

[0013] When the acquisition card system is in working state, the input capture IO port of MCU is in the state of receiving data. At this time, the voltage comparator circuit will send the state of each node to MCU. When the pressure reaches the threshold set by the potentiometer, the voltage comparator will output a high level, and MCU will mark the state of the point as "1", that is, it is considered that the node has pressure, otherwise it will mark the state of the point as "0", that is, it is considered that the node has no pressure. This method can avoid the work of ADC, thereby saving a lot of MCU running time, shortening the cycle of each scan, and improving the speed of detection. At the same time, it also basically meets the system's requirements for data measurement. When the MCU collects the pressure data of each node, it can roughly draw a pressure distribution map of the passenger's body, thereby judging the current sitting posture of the person.

[0014] In order to further improve the efficiency of detection, the present invention proposes a fast detection method for scanning the edge of the pressure distribution map. The patent of the present invention provides a method for fast detection in clockwise direction, and only provides tools for clockwise edge detection. The principle of counterclockwise direction is the same and will not be repeated. The specific detection method is as follows:

[0015] Step 1: Preparation

[0016] For a sensor array with m rows and n columns, two rows and two columns of virtual sensors are virtualized around it. The virtual sensor rows and columns do not correspond to the actual sensors. Accordingly, when the sensors in the virtual rows and columns are detected, the actual circuit does not perform any operation. Therefore, the result obtained by the detection is always the initial value "0" of the circuit. Therefore, the significance of the virtual sensor is only for the convenience of subsequent operations. Then establish the coordinate system of the sensor array, with its Y axis pointing to the front of the wheelchair occupant and the X axis pointing from the left of the wheelchair occupant to the right of the wheelchair occupant; with the coordinate system, the sensor in the yth row and xth column can be marked as S(x, y);

[0017] Create a bit variable matrix S to record the measured values ​​of sensors with m rows and n columns. The initial value of S is all 0.

[0018]

[0019] Establish a circular shift register P, whose initial value is: P = [1 2 3 4 3 2]. If the shift register P is shifted left once, its value becomes: P = [2 3 4 3 2 1]. If it is shifted left again, its value becomes: P = [3 4 3 2 1 2], and so on. The first value of the circular shift register P is used to indicate the edge detection tool to be selected. The patent of this invention has four clockwise edge detections, and their numbers and meanings are shown in the following table. In the table, the center square of the 9-square grid marked with cross-hatching is the current sensor position, and its position and measurement value are Sc (x0, y0) = 1.

[0020]

[0021] Step 2: Select the initial detection sensor. Take x0 = (m + 1) / 2, y0 = (n + 1) / 2, and round them to integers (x0, y0). The sensor corresponding to this coordinate is recorded as Sc (x0, y0).

[0022] Step 3: Enter the sensor area that bears the weight of the human body. Detect the value of the sensor Sc(x0-1, y0). If Sc(x0-1, y0) = 0, it means that the sensor does not bear the weight of the human body, then record x0 = x0-1, y0 = y0, and repeat step 3; if Sc(x0-1, y0) = 1, it means that the sensor bears the weight of the human body, then record x0 = x0-1, y0 = y0, S(x0, y0) = 1, and enter step 4.

[0023] Step 4: Find the sensor with measured values ​​on the leftmost side of the same row in the load-bearing area. Detect the value of sensor Sc(x0-1, y0). If Sc(x0-1, y0) = 1, it means that this sensor bears the weight of the human body. Then record x0 = x0-1, y0 = y0, S(x0, y0) = 1, and repeat step 4. If Sc(x0-1, y0) = 0, it means that this sensor does not bear the weight of the human body. The sensor at the edge of the load-bearing area has been found. Then record the coordinates of the current edge load-bearing sensor S0 as (Sx0, Sy0) and proceed to step 5.

[0024] Step 5: Initialize the circular shift register. Since the above steps detect the left edge of the figure from the center to the left, the value of the circular shift register is: P = [1 2 3 4 3 2], that is, the edge detection tool with sequence number 1 should be selected.

[0025] Step 6. Select the edge detection tool according to the indication of the circular shift register and start edge detection.

[0026] Step 7. Detect the value of sensor Sc1. If Sc1=1, it means that this sensor bears the weight of the human body. Then assign the coordinates of Sc1 to the current sensor, that is: (x0, y0)=Scxy1, and record S(x0, y0)=1, and proceed to step 11; if Sc1=0, proceed to step 8.

[0027] Step 8: Detect the value of sensor Sc2. If Sc2=1, it means that the sensor bears the weight of the human body. Then assign the coordinates of Sc2 to the current sensor, that is, (x0, y0)=Scxy2, and record S(x0, y0)=1. Go to step 11. If Sc2=0, go to step 9.

[0028] The ninth step is to detect the value of sensor Sc3. If Sc3=1, it means that this sensor bears the weight of the human body. Then the coordinates of Sc3 are assigned to the current sensor, that is: (x0, y0)=Scxy3, and S(x0, y0)=1, and proceed to the eleventh step; if Sc3=0, proceed to the tenth step.

[0029] Step 10: Shift the ring register one position to the left. The shift means changing the detection tool according to the rules. Return to step 6.

[0030] Step 11: Check if edge detection is complete

[0031] That is, check whether the binary vector (x0, y0) is equal to (Sx0, Sy0). If they are equal, it means that the edge detection is completed and enters the twelfth step; if they are not equal, return to the sixth step.

[0032] Step 12: Prepare to perfect the S matrix

[0033] The purpose of perfecting the S matrix is ​​to evaluate the sensor units in the graph that have not been tested by the circuit, and fill "1" in the corresponding positions. Since the calculation speed of the microprocessor is much faster than the speed of detecting the sensor through the circuit, the detection speed can be further accelerated. Its physical basis is to measure the distribution of gravity on the arm of the human body in a sitting position through a sensor, and the result is a continuous, single-connected and envelopable graph in the shape of a "concave" character. The S matrix can be perfected row by row or column by column. In the present application, the method of perfecting the S matrix by row and by column is the same. Taking the perfection of the S matrix by row as an example, set the starting column mark counter xs and the ending column mark counter xe, xs and xe are positive integers, and the maximum value is m+1. And set x=0, y=0.

[0034] Step 13: Find the row edge of the drawing. If x = m + 1, then jump to step 17; check S (x, y), if S (x, y) = 1, then xs = x, xe = x, and go to step 14; otherwise x = x + 1, repeat step 13.

[0035] Step 14: Find the end column that has been set to 1 by the circuit detection in a row of the S matrix, and find a starting point to fill "1": if x = m + 1, then jump to step 17. Check S (x + 1, y), if S (x + 1, y) = 1, then x = x + 1, xs = x, xe = x, repeat step 14; otherwise: x = x + 1, xe = x, go to step 15.

[0036] Step 15: Find the end column of the graph in the row that is not set to 1. If x = m + 1, then jump to step 15. Check S (x + 1, y), if S (x + 1, y) = 0, then x = x + 1, xe = x, repeat step 15; otherwise: x = x + 1, go to step 16

[0037] Step 16: Set all elements from S(xs, y) to S(xe, y) to 1.

[0038] Step 17: If x=m+1, it means that the row has been filled and the process goes to step 18; otherwise, the process goes back to step 13.

[0039] Step 18: Determine whether y=n+1. If yes, it means that the S matrix has been processed and go to step 19; otherwise, y=y+1 and return to step 13.

[0040] Step 19: Calculate the coordinates of the center of gravity (xf, yf) of the human body's weight distribution on the arm when sitting.

[0041]

[0042] At this point, all processes are completed.

[0043] The beneficial effects of the present invention are:

[0044] The device and method for precise and rapid detection of the center of gravity of the body posture of a wheelchair occupant described in the present invention use several sensors with relatively small areas to measure the center of gravity of the human body in a sitting position, and the measurement error is no greater than the geometric size of each measuring unit. In addition, in order to achieve rapid data acquisition, the sensor acquisition card used in the present invention converts the analog value of the voltage into a digital value through a voltage comparator circuit, so that it can quickly determine whether pressure is applied to the node according to the set threshold. And a scanning algorithm for the contour of the pressure area is designed to avoid scanning a large number of invalid nodes, thereby improving the efficiency of sensor scanning detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a connection diagram of the thin film pressure sensor and acquisition card system.

[0046] Figure 2 This is the schematic diagram of the voltage comparator.

[0047] Figure 3 It is a schematic diagram of the pressure distribution when an occupant's right leg is placed on his left leg; the shaded area indicates that the sensor's measurement value is "1", and the blank area indicates that the measurement value is "0".

[0048] Figure 4 This is a schematic diagram of the sensor locations and nodes described in the first step.

[0049] Figure 5 A step diagram of a specific implementation method. DETAILED DESCRIPTION

[0050] The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0051] Take the 32×31 sensor array pattern as an example, Figure 3 The concave curve in the figure is a typical curve measured in practice. The shaded squares are sensors with a measurement value of 1, surrounded by the concave curve. The remaining squares represent sensors with a measurement value of 0. The black squares are nodes for changing detection tools during edge scanning. The detailed edge detection process is as follows:

[0052] According to the second step of the detection method of this patent, the coordinates of the starting node are calculated to be (16, 15.5) and rounded to (16, 16), that is, the sensor at the 16th row and 16th column is taken, recorded as Sc(16, 16).

[0053] According to the third step of the detection method of this patent, since Sc(16, 16)=1, the sensor area that bears the weight of the human body has been entered.

[0054] According to the fourth step of the detection method of the present invention, the edge node 1 on the concave curve is found, which is Sc(16, 6), and a total of 10 detections are performed.

[0055] According to the fifth and sixth steps of the detection method of this patent, node 1 is a left edge node, so the initial value of the ring register is: P = [1 2 3 4 3 2].

[0056] According to the seventh to ninth steps of the detection method of this patent, the edge node 2 is detected by tool 1, and a total of 29 detections are performed, of which two are used to detect the applicable conditions of the tool. There are three virtual sensors above node 2, and their values ​​must be 0. Therefore, after three detection cycles, it is found that edge detection tool 1 is no longer applicable. According to the rule, the ring register P is shifted left, then P = [2 3 4 3 2 1], tool 2 should be selected.

[0057] After 4 tests, we reach node 3, and tool 2 is no longer applicable. According to the rules, we shift the ring register P to the left, and then P = [3 43 2 1 2], so tool 3 should be selected.

[0058] After three tests, it is found that tool 3 is no longer applicable. According to the rules, the ring register P is shifted left, then P = [4 3 2 1 23], so tool 4 should be used.

[0059] After 35 tests, we reach node 4 and find that tool 4 is no longer applicable. According to the rules, we shift the ring register P to the left, then P = [3 2 1 2 3 4], so tool 3 should be selected.

[0060] Use tool 3, after 2 tests, it reaches node 5. The applicable conditions of tool 3 are no longer met. According to the rules, shift the ring register P to the left, then P = [2 1 2 3 4 3], so tool 2 should be used.

[0061] Use tool 2, after one test, it reaches node 6, and the applicable conditions of tool 2 are no longer met. According to the rules, shift the ring register P to the left, then P = [1 2 3 4 3 2], so tool 1 should be selected.

[0062] It is found that the applicable conditions of tool 1 are no longer valid. According to the rules, the ring register P is shifted left, then P = [2 3 4 3 21], and tool 2 should be used.

[0063] It is found that the applicable conditions of tool 2 are no longer valid. According to the rules, the ring register P is shifted left, then P = [3 4 3 2 12], and tool 3 should be used.

[0064] Using tool 3, after 21 tests, we reach node 7. The applicable conditions of tool 3 are no longer met. According to the rules, we shift the ring register P to the left, then P = [4 3 2 1 2 3], so tool 4 should be used.

[0065] Using tool 4, after 10 tests, we reach node 8. The applicable conditions of tool 4 are no longer met. According to the rules, we shift the ring register P to the left, then P = [3 2 1 2 3 4], so tool 3 should be used.

[0066] It is found that the applicable conditions of tool 3 are no longer valid. According to the rules, the ring register P is shifted left, then P = [2 1 2 3 43], and tool 2 should be used.

[0067] It is found that the applicable conditions of tool 2 are no longer valid. According to the rules, the ring register P is shifted left, then P = [1 2 3 4 32], and tool 1 should be selected.

[0068] Using tool 1, after 18 tests, it reaches node 9. The applicable conditions of tool 1 are no longer met. According to the rules, the ring register P is shifted left, then P = [2 3 4 3 2 1], so tool 2 should be used.

[0069] It is found that the applicable conditions of tool 2 are no longer valid. According to the rules, the ring register P is shifted left, then P = [3 4 3 2 12], and tool 3 should be used.

[0070] It is found that the applicable conditions of tool 3 are no longer valid. According to the rules, the ring register P is shifted left, then P = [4 3 2 1 23], and tool 4 should be used.

[0071] Using tool 4, after 12 tests, it reaches node 10. The applicable conditions of tool 4 are no longer met. According to the rules, the ring register P is shifted left, then P = [3 2 1 2 3 4], so tool 3 should be used.

[0072] Using tool 3, after 41 detections, we reach node 11. The applicable conditions of tool 4 are no longer met. According to the rules, we shift the ring register P to the left, then P = [2 1 2 3 4 3], so tool 2 should be used.

[0073] Using tool 2, after 41 tests, it reaches node 12. The applicable conditions of tool 2 are no longer met. According to the rules, the ring register P is shifted left, then P = [1 2 3 4 3 2], so tool 1 should be selected.

[0074] Using tool 1, after 5 detections, it is found to coincide with node 1, and the entire detection process is completed.

[0075] Correspondingly, the value of the sensor matrix S is:

[0076]

Claims

1. A precise and rapid detection device for the center of gravity of a wheelchair occupant's body posture, characterized by: It includes several thin film pressure sensors and corresponding acquisition card systems, wherein the thin film pressure sensors are arranged at the seat cushion and backrest of the wheelchair, the acquisition card system includes an MCU controller and an acquisition card circuit, the thin film pressure sensors are connected to the MCU controller through a control line, and a voltage comparator circuit is arranged in the acquisition card circuit; The method for accurately and quickly detecting the center of gravity of the wheelchair occupant's body posture comprises the following steps: Take clockwise edge detection as an example: Step 1: Preparation For a sensor array with m rows and n columns, two rows and two columns of virtual sensors are virtualized around it. The virtual sensor rows and columns do not correspond to the actual sensors. Accordingly, when the sensors in the virtual rows and columns are detected, the actual circuit does not perform any operation. Therefore, the detection result is always the initial value "0" of the circuit. Therefore, the significance of virtual sensors is only for the convenience of subsequent operations. Then establish the coordinate system of the sensor array, with the Y axis pointing to the front of the wheelchair occupant and the X axis pointing from the left to the right of the wheelchair occupant; with the coordinate system, the sensor in the yth row and xth column can be marked as S(x, y); Create a bit variable matrix S to record the measured values ​​of sensors with m rows and n columns. The initial value of S is all 0. Establish a circular shift register P, whose initial value is: P = [123432]. If the shift register P is shifted left once, its value becomes: P = [234321]. If it is shifted left again, its value becomes: P = [343212], and so on. The first value of the circular shift register P is used to indicate the edge detection tool to be selected. This method has four clockwise edge detections, and their positions and measurements are Sc (x0, y0) = 1. Step 2: Select the initial detection sensor Take x0 = (m + 1) / 2, y0 = (n + 1) / 2, and round them to integers (x0, y0). The sensor corresponding to this coordinate is recorded as Sc (x0, y0); Step 3: Enter the sensor area that bears the weight of the human body Detect the value of sensor Sc(x0-1, y0). If Sc(x0-1, y0)=0, it means that the sensor does not bear the weight of the human body, then record x0=x0-1, y0=y0, and repeat the third step; if Sc(x0-1, y0)=1, it means that the sensor bears the weight of the human body, then record x0=x0-1, y0=y0, S(x0, y0)=1, and go to the fourth step; Step 4: Find the sensor with measured values ​​on the far left of the same row in the load area Detect the value of sensor Sc(x0-1, y0). If Sc(x0-1, y0)=1, it means that the sensor bears the weight of the human body. Then record x0=x0-1, y0=y0, S(x0, y0)=1, and repeat step 4. If Sc(x0-1, y0)=0, it means that the sensor does not bear the weight of the human body. The sensor at the edge of the load-bearing area has been found. Then record the coordinates of the current edge load-bearing sensor S0 as (Sx0, Sy0), and go to step 5. Step 5: Initialize the ring shift register Since the above steps detect the left edge of the figure from the center to the left, the value of the circular shift register is: P = [123432], that is, the edge detection tool No. 1 should be selected Step 6: Select the edge detection tool according to the indication of the circular shift register and start edge detection; Step 7: Detect the value of sensor Sc1. If Sc1=1, it means that the sensor bears the weight of the human body. Then assign the coordinates of Sc1 to the current sensor, that is, (x0, y0)=Scxy1, and record S(x0, y0)=1, and proceed to step 11; if Sc1=0, proceed to step 8; Step 8: Detect the value of sensor Sc2. If Sc2=1, it means that the sensor bears the weight of the human body. Then assign the coordinates of Sc2 to the current sensor, that is, (x0, y0)=Scxy2, and record S(x0, y0)=1; go to step 11; if Sc2=0, go to step 9; Step 9, detect the value of sensor Sc3. If Sc3=1, it means that this sensor bears the weight of the human body, then assign the coordinates of Sc3 to the current sensor, that is: (x0, y0)=Scxy3, and record S(x0, y0)=1, and go to step 11; if Sc3=0, go to step 10; Step 10: Shift the ring register left by one position, and the shift means changing the detection tool according to the rule; return to step 6; Step 11: Check if edge detection is complete That is, check whether the binary vector (x0, y0) is equal to (Sx0, Sy0). If they are equal, it means that the edge detection is completed and the process goes to step 12. If they are not equal, the process goes back to step 6. Step 12: Prepare to perfect the S matrix The purpose of improving the S matrix is ​​to evaluate the sensor units in the graph that have not been tested by the circuit, and fill "1" in the corresponding positions; improve the S matrix row by row or column by column, set the starting column mark counter xs, and the ending column mark counter xe, xs, xe are positive integers, and the maximum value is m+1; and set x=0, y=0; Step 13: Find the row edge of the drawing; if x = m + 1, then jump to step 17; check S (x, y), if S (x, y) = 1, then xs = x, xe = x, go to step 14; otherwise x = x + 1, repeat step 13; Step 14: Find the end column that has been set to 1 by circuit detection in a row of the S matrix, and find a starting point to fill with "1": if x = m + 1, then jump to step 17; check S (x + 1, y), if S (x + 1, y) = 1, then x = x + 1, xs = x, xe = x, repeat step 14; otherwise: x = x + 1, xe = x, go to step 15; Step 15: Find the end column of the graphic in the row that is not set to 1 If x=m+1, then jump to step 15; check S(x+1, y), if S(x+1, y)=0, then x=x+1, xe=x, repeat step 15; otherwise: x=x+1, go to step 16; Step 16: Set all elements from S(xs, y) to S(xe, y) to 1; Step 17: If x=m+1, it means that the row has been filled, and go to step 18; otherwise, return to step 13; Step 18: Determine if y=n+1. If yes, it means that the S matrix has been processed and the process goes to step 19. Otherwise, y=y+1 and the process goes back to step 13. Step 19: Calculate the coordinates of the center of gravity (xf, yf) of the human body's gravity distribution on the arm in the sitting position; At this point, all processes are completed.

2. The precise and rapid detection device for the center of gravity of a wheelchair occupant's body posture according to claim 1 is characterized by: The number of acquisition nodes of the thin film pressure sensor is (24-32)×(24-32).

3. The precise and rapid detection device for the center of gravity of a wheelchair occupant's body posture according to claim 1 is characterized in that: The control line is provided with a control electronic switch.

4. The precise and rapid detection device for the center of gravity of a wheelchair occupant's body posture according to claim 1 is characterized in that: The voltage comparator circuit is composed of a voltage comparator chip and a potentiometer. The potentiometer is used to adjust the set voltage threshold. The input voltage from the operational amplifier is compared with the threshold voltage. When the threshold voltage is reached, the voltage comparator circuit outputs a high level, otherwise it outputs a low level. The output signal of the voltage comparator circuit is connected to the IO port of the MCU controller to collect the status information of each node.

5. The precise and rapid detection device for the center of gravity of a wheelchair occupant's body posture according to claim 1 is characterized in that: When the acquisition card system is in working state, the input capture IO port of MCU is in the state of receiving data; at this time, the voltage comparator circuit will send the state of each node to MCU. When the pressure reaches the threshold set by the potentiometer, the voltage comparator will output a high level, and MCU will mark the state of the point as "1", that is, it is considered that the node has pressure, otherwise it will mark the state of the point as "0", that is, it is considered that the node has no pressure; after MCU collects the pressure data of each node, it draws a pressure distribution map of the passenger's body, thereby judging the current sitting posture of the person.

Images