Correction method

Abstract

A correction method is disclosed for a virtual reality system, which maps a real hand pose of a user to a virtual hand pose. A computer of the virtual reality system multiplies instant detection results of a plurality of first sensors of a finger pose detection device of the virtual reality system, which are respectively located on a plurality of fingers of a hand of the user, by a first rotation matrix and a second rotation matrix to calculate a plurality of virtual finger poses of the virtual hand pose, wherein each of the first sensors and the second sensors comprises an inertial measurement unit and an element sensing unit, the first rotation matrix is a conversion matrix between a coordinate system of each of the first sensors and a coordinate system of a second sensor of the finger pose detection device which is located on a palm of the hand, and the second rotation matrix is a conversion matrix between the coordinate system of the second sensor and a coordinate system of a control handle of the virtual reality system. The computer directly uses instant detection results of the control handle as a virtual palm pose of the virtual hand pose.

Description

[0001] This invention is a divisional application of application number 201910695567.8 filed on July 30, 2019, entitled "Finger Posture Detection Device, Control Assembly and Correction Method". Technical Field

[0002] This invention relates to a finger posture detection device and method, and more particularly to a finger posture detection device, control assembly and correction method suitable for virtual reality. Background Technology

[0003] Virtual Reality (VR) technology is becoming increasingly popular, and the most common input device is the controller. However, traditional controllers can only provide simple input functions such as clicking or selecting objects, and cannot detect the degree of finger bending. Therefore, they cannot be used in virtual environments that require finger posture, such as playing the piano or typing on a keyboard. Summary of the Invention

[0004] This invention provides a finger posture detection device that can detect the user's finger state.

[0005] The present invention relates to a finger posture detection device suitable for a control handle. The control handle includes a grip portion adapted to be held by a hand and moved in three-dimensional space. The finger posture detection device includes a first wear portion, a plurality of second wear portions, and a plurality of first sensors. The first wear portion is adapted to detachably surround the grip portion of the control handle. When the grip portion is connected to the first wear portion, the grip portion is positioned on one side of the first wear portion. Each of the second wear portions is independently connected to the first wear portion. When a force is applied to one of the second wear portions, the corresponding second wear portion moves toward the grip portion. The first sensors are respectively disposed on the second wear portions for detecting the position or movement of the second wear portion relative to the grip portion.

[0006] In one embodiment of the present invention, the aforementioned plurality of second wearable portions are respectively worn on a plurality of fingers of the hand. Each finger includes a first phalanx, a second phalanx, and a third phalanx sequentially away from a fingertip. The second wearable portion is worn on the third phalanx.

[0007] In one embodiment of the present invention, the control handle described above includes a plurality of sensing elements. The sensing elements are disposed on the grip portion, and the sensing elements and the first sensor sense each other to generate a signal.

[0008] In one embodiment of the present invention, the sensing element includes a magnet, and the sensing element is alternately oriented toward the first sensor with its S magnetic pole or N magnetic pole respectively.

[0009] In one embodiment of the present invention, the finger posture detection device further includes a second sensing element. The second sensing element extends toward the first sensor and is disposed in the first wearable portion to sense each other with the first sensor to generate a signal. The second wearable portion is disposed between the second sensing element and the grip portion.

[0010] In one embodiment of the present invention, the finger posture detection device further includes a control module and a second sensor. The control module is disposed on the first wearable part. The signals sensed by the first sensor are combined with the signals sensed by the second sensor to calculate the position of each first sensor relative to the gripping part.

[0011] In one embodiment of the present invention, when the first sensor detects the second sensing element, the signal detected by the first sensor is corrected.

[0012] In one embodiment of the present invention, the first sensor includes an inertial measurement unit and a component sensing unit. When the component sensing unit detects the second sensing element, the signal detected by the inertial measurement unit corresponding to the relative positions of the second wearable part and the gripping part is corrected.

[0013] The present invention relates to a finger posture detection device suitable for a control handle. The control handle includes a grip portion adapted to be held by a hand and moved in three-dimensional space. The grip portion is provided with multiple sensors. The finger posture detection device includes a first wear portion, multiple second wear portions, and multiple sensing elements. The first wear portion is adapted to detachably surround the grip portion of the control handle. When the grip portion is connected to the first wear portion, the grip portion is positioned on one side of the first wear portion. Each of the second wear portions is independently connected to the first wear portion. When a force is applied to one of the second wear portions, the corresponding second wear portion moves toward the grip portion. Sensing elements are respectively disposed on the second wear portions, providing sensors for corresponding sensing elements to detect the position or movement of the second wear portion relative to the grip portion.

[0014] In one embodiment of the invention, the second wearing portion is worn on multiple fingers of the hand. Each finger includes a first phalanx, a second phalanx, and a third phalanx sequentially away from a fingertip. The second wearing portion is worn on the third phalanx.

[0015] In one embodiment of the invention, the sensing element includes a magnet. The sensing element alternately faces the grip portion with its S pole or N pole facing it.

[0016] The control assembly of the present invention is applied to virtual reality and includes a control handle and the aforementioned finger posture detection device.

[0017] In one embodiment of the present invention, the control handle includes a sensing element, and the sensing element is disposed on the grip portion.

[0018] The correction method of this invention is applicable to a virtual reality system. The virtual reality system maps a user's real hand posture to a virtual hand posture. The virtual hand posture includes multiple virtual finger postures and a virtual palm posture. The virtual reality system includes a control handle for being held by the user's hand, a finger posture detection device detachably connected to the control handle and worn on the hand, and a computer. The finger posture detection device includes multiple first sensors and a second accessory sensor. The first sensors are located on multiple fingers of the hand. The second accessory sensor is located on the palm of the hand. The first sensors include an inertial measurement unit (IMU) and a component sensing unit. The second accessory sensor includes an inertial measurement unit and a component sensing unit. The correction method includes the following steps: The computer multiplies the real-time detection results of each first sensor by a first rotation matrix and a second rotation matrix to calculate the virtual finger posture. The first rotation matrix is ​​a transformation matrix between the coordinate systems of the first sensors and the second sensors. The second rotation matrix is ​​a transformation matrix between the coordinate systems of the second sensors and the coordinate system of the control handle. The computer directly uses the real-time detection results of the control handle as the virtual hand gesture.

[0019] In one embodiment of the present invention, the above-mentioned correction method further includes correcting the virtual finger posture of the virtual hand posture to the point that the fingers are straight and close together relative to the palm when the difference between the instantaneous readings of the accelerometers of each first sensor and the instantaneous readings of the accelerometers of the second sensor is less than a preset value, and the difference between the instantaneous readings of the element sensing units of adjacent first sensors is less than another preset value.

[0020] The correction method of the present invention is applicable to a virtual reality system. The virtual reality system includes a control handle for being held by a user's hand, a finger posture detection device detachably connected to the control handle and worn on the hand, and a computer. This correction method includes the following steps: The computer calculates an actual bending range of a finger of the user's hand relative to a palm based on initial detection results from the finger posture detection device. The actual bending range has an upper boundary and a lower boundary. The computer maps a virtual bending range of a virtual finger corresponding to the actual bending range. When the degree of finger bending calculated based on the real-time detection results of the finger posture detection device exceeds the upper boundary of the actual bending range, the upper boundary of the actual bending range is updated by the degree of bending exceeded. When the degree of finger bending calculated based on the real-time detection results of the finger posture detection device exceeds the lower boundary of the actual bending range, the lower boundary of the actual bending range is updated by the degree of bending exceeded.

[0021] Based on the above, the finger posture detection device of the present invention can detect the posture of the corresponding finger through a first sensor configured in the second wearable part, thereby enabling more complex input functions in virtual reality. The correction method of the present invention, through switching between the first sensor, the second accessory sensor, and the control handle, can enable a computer to correct the virtual hand posture in real time based on the user's actual hand posture. Furthermore, the computer can update the upper and lower boundaries of the actual bending range of the finger based on the real-time detection results of the finger posture detection device, and then estimate and correct the virtual hand posture based on the new bending range.

[0022] To make the above features and advantages of the present invention more apparent and understandable, specific embodiments are described below in conjunction with the accompanying drawings. Attached Figure Description

[0023] Figure 1A This is a schematic diagram of a finger posture detection device worn on a user's hand according to an embodiment of the present invention;

[0024] Figure 1B yes Figure 1A An illustration of the user's hand is omitted.

[0025] Figure 1C yes Figure 1A A planar schematic diagram of a finger posture detection device worn on a user's hand;

[0026] Figure 1D yes Figure 1A A schematic diagram of the assembly of the finger posture detection device and the control handle;

[0027] Figure 1E It is used to include Figure 1A A flowchart of a correction method for a virtual reality system using a finger posture detection device;

[0028] Figure 2 This is a diagram of the knuckles of the fingers;

[0029] Figure 3 This is a schematic diagram of a finger posture detection device according to another embodiment of the present invention, applicable to a control handle;

[0030] Figure 4 This is a schematic diagram of a finger posture detection device according to another embodiment of the present invention, applicable to a control handle;

[0031] Figure 5A This is a schematic diagram of a finger posture detection device worn on a user's hand according to another embodiment of the present invention;

[0032] Figure 5B It is used to include Figure 5AA flowchart of a correction method for a virtual reality system using a finger posture detection device;

[0033] Figure 6 This is a schematic diagram of a finger posture detection device according to another embodiment of the present invention, applicable to a control handle.

[0034] Symbol Explanation

[0035] 50: Control handle

[0036] 52: Grip section

[0037] 54: First sensing element

[0038] 54b: Sensor

[0039] 100, 100a, 100b, 100c, 100d: Finger posture detection device

[0040] 110: First Wearable Department

[0041] 112: Sleeve

[0042] 120: Second Wearing Department

[0043] 130, 130a, 130c, 130d: First sensor

[0044] 130b: Sensing element

[0045] 140: Second sensor

[0046] 140a: Second sensing element

[0047] 150: Control Module

[0048] 160: Sheet-like extension

[0049] 170: Elastic rope

[0050] 171: Tightening mechanism

[0051] 172: Dual-shaft mechanism

[0052] 173: Flexible elements

[0053] K1: First knuckle

[0054] K2: Second knuckle

[0055] K3: Third knuckle Detailed Implementation

[0056] Figure 1A This is a schematic diagram of a finger posture detection device worn on a user's hand, according to an embodiment of the present invention. Figure 1B yes Figure 1A An illustration of the user's hand is omitted. Figure 1C yes Figure 1A A schematic diagram of a finger posture detection device worn on the user's hand. Please refer to... Figure 1A , Figure 1B and Figure 1C The finger posture detection device in this embodiment is applied to a virtual reality system. The virtual reality system includes a control handle 50 and a computer. The control handle 50 is adapted to be worn on a user's hand and can be connected to the computer via wired or wireless means to exchange signals with the computer. In addition, the virtual reality system also includes a head-mounted display, which is adapted to be worn on a user's head and can also be connected to the computer via wired or wireless means to exchange signals with the computer.

[0057] In this embodiment, the control handle 50 includes a grip portion 52, adapted to be held by a user's hand and moved in three-dimensional space. The finger posture detection device 100 of this embodiment includes a first wearable portion 110, a plurality of second wearable portions 120, and a plurality of first sensors 130. The first wearable portion 110 is adapted to be detachably connected to and surround the grip portion 52 of the control handle 50. When the grip portion 52 is connected to the first wearable portion 110, the grip portion 52 is positioned on one side of the first wearable portion 110. Each of the second wearable portions 120 is independently connected to the first wearable portion 110. When a force is applied to one of the second wearable portions 120, the corresponding second wearable portion 120 moves toward the grip portion 52. The first sensors 130 are respectively disposed on the second wearable portions 120 for detecting the position or movement of the second wearable portions 120 relative to the grip portion 52. When a user wears the finger posture detection device 100 of this embodiment, the first wearing part 110 is positioned on the user's palm, and the second wearing parts 120 are respectively worn on multiple fingers of the user.

[0058] Figure 2 This is a diagram of the knuckles of the fingers. Please refer to it. Figure 1A and Figure 2 In this embodiment, each finger includes a first phalanx K1, a second phalanx K2, and a third phalanx K3 sequentially located away from a fingertip. In this embodiment, these second wearable parts 120 are worn on multiple fingers respectively, and correspond to the third phalanx K3 of each finger.

[0059] To more clearly illustrate the finger posture detection device 100 of this embodiment, Figure 1B The user's hands are omitted from the drawing. Please refer to [reference needed]. Figure 1A and Figure 1BIn this embodiment, the first sensor 130 includes an inertial measurement unit (IMU) and a component sensing unit (e.g., an electronic compass). In other embodiments, the first sensor may also include a bending sensor, a tension sensor, or a magnetic element, and the present invention is not limited thereto.

[0060] In this embodiment, the first sensor 130 is disposed on the side of the second wearable parts 120 away from the gripping part 52. That is, when the user wears the finger posture detection device 100, the position of the first sensor 130 corresponds to the outer side of the third phalanx K3 of each individual finger. During the process of the user wearing the finger posture detection device 100, the first sensor 130 can detect the degree of bending (posture) of the third phalanx K3 of each finger. By estimating the finger joint movement, the posture of each finger can be deduced (i.e., the degree of bending of the first phalanx K1 and the second phalanx K2 can be calculated). It should be noted that the principle of the finger joint movement is that when the finger bends from the open state towards the palm, the joints between each phalanx will rotate inward, and the joints will rotate in the same direction. The relationship between the rotation angle of individual phalanxes and the total rotation angle can be obtained statistically. With this statistical result, the rotation of the first phalanx K1 and the second phalanx K2 can be calculated by evaluating the amount of rotation of the third phalanx K3 joint. Of course, in other embodiments, the first sensor 130 can also be configured in the first phalanx K1 or the second phalanx K2, and the posture of each finger can also be derived by estimating the movement of the finger joints.

[0061] In this embodiment, as Figure 1C As shown, the finger posture detection device 100 further includes a control module 150 and a second sensor 140. The control module 150 is disposed on the first wearable part 110. The second sensor 140 includes an inertial measurement unit (IMU) and a component sensing unit (e.g., an electronic compass), and the second sensor 140 is disposed on the first wearable part 110 and electrically connected to the control module 150. The signals sensed by the first sensor 130 are combined with the signals sensed by the second sensor 140 to calculate the position of each first sensor 130 relative to the grip part 52.

[0062] like Figure 1BAs shown, in this embodiment, the control handle 50 may further include a first sensing element 54. The first sensing element 54 is, for example, a capacitive sensor or a pressure sensor, and is disposed on the grip portion 52. In other embodiments, the first sensing element may also be a Hall sensor, a magnetic element, a proximity sensor, or a miniature ultrasonic sensor, and the present invention is not limited thereto. In this embodiment, the first sensor 130 and the first sensing element 54 sense each other to generate a signal. When the first sensor 130 includes an inertial measurement unit, the inertial measurement unit itself will generate errors due to the accumulation of time during operation, causing the data (i.e., the detection result) of the inertial measurement unit to gradually become distorted. In this embodiment, by means of a first sensing element 54 configured on the control handle 50, when any finger of the user touches the corresponding position (i.e., the first sensing element 54) on the control handle 50, the inertial measurement unit of the first sensor 130 on the finger is automatically calibrated. This allows the inertial measurement unit of the first sensor 130 to be quickly repositioned and calibrated, avoiding distortion caused by accumulated errors in the inertial measurement unit of the first sensor 130 due to prolonged operation. In addition, besides being used to calibrate the inertial measurement unit in real time, the first sensing element 54 can also provide the user with a richer operating experience. For example, pressing the corresponding first sensing element 54 (e.g., a pressure sensor) can serve as a mechanism to trigger another event.

[0063] Figure 1D yes Figure 1A A schematic diagram showing the assembly of the finger posture detection device and the control handle. Please refer to the diagram. Figure 1A and Figure 1D The control handle 50 is another form. The finger posture detection device 100 includes a body 112 conforming to the shape of the grip portion 52 of the control handle 50, which can be fitted onto the grip portion 52. Both ends of the first wearing portion 110 are flexibly connected to the body 112, allowing the first wearing portion 110 to be detachably connected to the control handle 50 through the fitting between the body 112 and the grip portion 52. In this case, the first sensing element 54 (e.g., a magnet, Hall sensor, capacitive sensor, or pressure sensor) can be repositioned inside the body 112. One end of the first wearing portion 110 is connected to the body 112 in an adjustable length manner, allowing the user's palm to be bound between the first wearing portion of the finger posture detection device 100 and the grip portion 52 of the control handle 50, ensuring that the control handle 50 does not detach from the user's palm when the user releases their hand. Applying the finger posture detection device 100 of this embodiment to the control handle 50 expands the finger detection function of this control handle 50.

[0064] In this embodiment, as Figure 1DAs shown, the lower end of the first wearable part 110 is connected to the sleeve 112 via an elastic cord 170 and a tightening mechanism 171 through which the elastic cord 170 passes, allowing the user to adjust the tightness according to the size of their hand. The tightening mechanism 171 may include a disc and an element having a trapezoidal channel for receiving the disc. In addition, the upper end of the first wearable part 110 can be connected to the sleeve 112 in sequence via a dual-axis mechanism 172 and a flexible element 173. The dual-axis mechanism 172 provides an additional rotation mechanism (e.g., dual rotational degrees of freedom) to adapt to different hand angles caused by different grips of the user, allowing the first wearable part 110 to fit more snugly on the back of the hand. Furthermore, the above-mentioned component configuration also greatly improves the comfort and stability of wearing the garment, preventing significant positional misalignment due to excessive user movements.

[0065] In this embodiment, the first wearable part 110 can be made of an elastic material. When the user wears the corresponding four fingers (excluding the thumb) into the second wearable part 120, the elastic first wearable part 110 can attach the control handle 50 to the hand, thus providing a secure grip. During the use of the finger posture detection device 100, the user does not need to hold the control handle 50 for a long time, and the fingers can freely perform other tasks, such as answering a phone call or picking up other objects without putting down the control handle 50, thus greatly increasing the user's hand freedom. Furthermore, these elastic second wearable parts 120 can be fixed to the user's third phalanges K3, so actuators (not shown) can be placed in each second wearable part 120 to provide individual tactile feedback for different fingers (i.e., only the fingers interacting with virtual objects will feel the corresponding tactile feedback). In other embodiments, other sensors, such as PPG (heart rate sensor), can also be embedded in the second wearable part 120. Since these second wearable parts 120 can be attached to the third phalanx K3 for a long time, the sensors in the second wearable parts 120 can effectively detect and record the user's physiological information.

[0066] Figure 1E It is used to include Figure 1A A flowchart illustrating a calibration method for a virtual reality system using a finger posture detection device. This virtual reality system maps a user's real hand posture (i.e., the user's real-time hand posture) to a virtual hand posture (i.e., the virtual character's real-time hand posture). The virtual hand posture includes multiple virtual finger postures and a virtual palm posture. The virtual palm posture is based on the posture of the control handle 50, while the virtual finger postures are based on the posture of the first sensor 130 (inertial measurement unit) on each individual finger. Please refer to... Figure 1A , Figure 1B and Figure 1E , Figure 1E The correction method shown includes the following steps. First, the computer records the initial synchronization position's pose and the virtual hand pose (these poses are mapped according to quaternions), and calculates the rotational relationship (i.e., the transformation matrix) between them. Since there is a rotational relationship between the pose of the control handle 50 and the pose of the first sensor 130, this rotational relationship must be calculated in advance to correct the virtual hand pose. Next, the first sensor 130 installed on each finger is scanned, and the coordinates of the first sensor 130 are aligned. The purpose of aligning the first sensor 130 is to ensure that the reference points of the first sensor 130 are the same during rotation, thereby unifying the relative poses of each first sensor 130. After entering the game, the computer determines whether the update is of the virtual palm pose (corresponding to the quaternion of the control handle) or the virtual finger pose (corresponding to the quaternion of the IMU), and calculates the first rotation matrix and the second rotation matrix accordingly. Specifically, the first rotation matrix is ​​the transformation matrix between the coordinate systems of individual first sensors 130 and the coordinate systems of the second sensor 140. The second rotation matrix is ​​the transformation matrix between the coordinate system of the second sensor 140 and the coordinate system of the control handle 50. When the posture to be updated is not a virtual hand, the posture of the inertial measurement unit is used to update the virtual finger posture. In this case, the computer multiplies the real-time detection results (i.e., posture matrix) of each first sensor 130 by the first rotation matrix and the second rotation matrix to calculate the posture of each virtual finger. Alternatively, when the posture to be updated is a virtual hand, the posture of the control handle is used directly as the posture of the virtual hand. Then, the computer updates the entire virtual hand posture based on the new virtual finger posture and the virtual hand posture (i.e., the user's real hand posture).

[0067] After updating the virtual hand pose, such as Figure 1E As shown, according to this correction method, the difference between the instantaneous reading of the accelerometer of each first sensor 130 and the accelerometer value of the second accessory sensor 140, as well as the difference between the instantaneous readings of the electronic compasses of two adjacent first sensors 130, can be calculated by computer. When the difference between the instantaneous readings of the accelerometers of each first sensor 130 and the instantaneous readings of the accelerometers of the second accessory sensor 140 is less than a preset value, and the difference between the instantaneous readings of the electronic compasses of adjacent first sensors 130 is less than another preset value, the computer will correct the virtual finger posture of the virtual hand gesture to make the fingers straight and together relative to the palm (e.g., ...). Figure 1C (As shown).

[0068] The above correction method can avoid the distortion of the virtual finger caused by accumulated errors in the inertial measurement unit due to excessively long operation time. Furthermore, through... Figure 1EThe correction method shown allows the computer to update the virtual finger posture based on the real finger posture simply by extending the four fingers at any angle, achieving an immediate correction effect.

[0069] Figure 3 This is a schematic diagram of a finger posture detection device according to another embodiment of the present invention, applicable to a control handle. Figure 3 The finger posture detection device 100a shown is in conjunction with Figure 1A , 1B Similar to the finger posture detection device 100 shown, the difference lies in that the first sensor 130a of the finger posture detection device 100a is, for example, a bending sensor or a stretching sensor. In this embodiment, the degree of bending of the third phalanx K3 of each finger is sensed using the first sensor 130a, which can detect bending or stretching. Utilizing the characteristics of the first sensor 130a, the degree of bending of the third phalanx K3 can be obtained, and subsequently, the posture of each finger can be deduced through estimation of finger joint movement.

[0070] Figure 4 This is a schematic diagram of a finger posture detection device according to another embodiment of the present invention, applicable to a control handle. Figure 4 The finger posture detection device 100b shown is in conjunction with Figure 1A , 1B Similar to the finger posture detection device 100 shown, the difference lies in that the grip portion 52 of this embodiment is provided with multiple sensors 54b, and the finger posture detection device 100b includes multiple sensing elements 130b. These sensing elements 130b are respectively disposed on the second wear portion 120 and correspond to the sensing elements 130b, for detecting the position or movement of the second wear portion 120 relative to the grip portion 52. When the finger posture detection device 100b is worn on the user's hand, the position of the sensing elements 130b corresponds to the inner side of the third phalanx K3 of each finger. In this embodiment, the sensors 54b are, for example, Hall sensors and are embedded in the grip portion 52. The sensors 54b and the sensing elements 130b sense each other to generate signals. Utilizing the characteristic that the Hall sensor can sense the strength of magnetic force, the conversion relationship between the degree of bending of each finger and the magnetic force sensing intensity can be obtained. By obtaining the degree of bending of the third phalanx K3 and estimating the finger joint movement, the posture of each finger is then deduced.

[0071] Furthermore, multiple sensing elements 130b within the second wearable portion 120 are arranged alternately. Each sensing element 130b includes a magnet, and these sensing elements 130b are alternately oriented with their S or N magnetic poles facing the grip portion 52, causing the magnetic polarities of adjacent sensing elements 130b to be interleaved. Because the magnetic field lines of the sensing elements 130b on adjacent fingers are in opposite directions, the signals sensed by the corresponding sensors 54b (Hall sensors) exhibit a difference between positive and negative directions, preventing sensing errors caused by finger misalignment. In this way, when a finger misalignment causes a non-corresponding Hall sensor to output a signal, the system can quickly identify the signal generated by the incorrect finger's proximity, thereby achieving a correction function.

[0072] Figure 5A This is a schematic diagram of a finger posture detection device worn on a user's hand, according to another embodiment of the present invention. Figure 5A The finger posture detection device 100c shown is in conjunction with Figure 1A , Figure 1B Similar to the finger posture detection device 100 shown, the difference lies in that the finger posture detection device 100c in this embodiment further includes a sheet-like extension 160 and a second sensing element 140a. The second sensing element 140a extends toward the first sensor 130c and is disposed in the first wearable part 110 to sense each other with the first sensor 130c to generate a signal. The second wearable part 120 is disposed between the second sensing element 140a and the gripping part (not shown). Figure 5A )between.

[0073] In this embodiment, when the user wears the finger posture detection device 100c, the sheet-like extension 160 is positioned close to the back of the user's hand and is suitable as a sensing reference position for the first sensor 130c. The first sensor 130c is, for example, a Hall sensor, and the second sensing element 140a is, for example, a magnetic element disposed on the sheet-like extension 160. When the first sensor 130c detects the second sensing element 140a, the signal detected by the first sensor 130c is corrected. That is, when the finger is bent or straightened, the first sensor 130c obtains different values ​​depending on the distance between the third phalanx K3 and the sheet-like extension 160. At the same time, by combining the degree of bending of the third phalanx K3 of each finger, the degree of bending of the first phalanx K1 and the second phalanx K2 is estimated, and finally the posture of each finger is obtained. In other embodiments, the first sensor may also be a proximity sensor or a miniature ultrasonic sensor, which uses light reflection or ultrasound to detect the distance between the first sensor and the second accessory sensor to obtain the distance relationship between each finger and the reference position, and then detect the bending posture of each finger.

[0074] In other embodiments, the first sensor may include an inertial measurement unit and an element sensing unit. When the element sensing unit detects the second sensing element, the signal detected by the inertial measurement unit corresponding to the relative positions of the second wearable part and the gripping part is corrected.

[0075] Figure 5B It is used to include Figure 5A A flowchart illustrating a calibration method for a virtual reality system using a finger posture detection device. Please refer to the following: Figure 5A and Figure 5B This correction method includes the following steps. First, the computer of the virtual reality system calculates an actual bending range (e.g., angle range) of the user's finger relative to the palm based on the initial detection results (e.g., magnetic force value, resistance value, capacitance value, or other data) of the first sensor 130c. This actual bending range has an upper bound (e.g., maximum angle value) and a lower bound (e.g., minimum angle value). Here, the upper bound is the maximum (or minimum) value sensed by the first sensor 130c when it is closest to the sheet-like extension 160, and the lower bound is the minimum (or maximum) value sensed by the first sensor 130c when it is furthest from the sheet-like extension 160. Next, the computer maps the actual bending range to the virtual bending range of the corresponding virtual finger. After entering the game, based on the interaction between the user and the virtual object, the computer continuously records the real-time detection results (e.g., magnetic force value, resistance value, capacitance value, or other data) of each finger. When the degree of finger bending calculated based on the real-time detection results exceeds the upper bound of the actual bending range, the upper bound of the actual bending range is updated with the degree of bending exceeding the upper bound. When the degree of finger bending calculated based on real-time detection results exceeds the lower bound of the actual bending range, the lower bound of the actual bending range is updated based on the degree of bending that exceeds the limit. Then, the computer uses the updated upper and lower bounds of the actual bending range as the estimated range of the finger bending posture, and thus synchronously corrects the virtual finger posture.

[0076] pass Figure 5B The calibration method shown can establish different detection ranges for different people, different hand shapes, and different gripping methods, thereby more accurately estimating and correcting the virtual finger posture in the virtual reality.

[0077] Figure 6 This is a schematic diagram of a finger posture detection device according to another embodiment of the present invention, applicable to a control handle. Figure 6 The finger posture detection device 100d shown is in conjunction with Figure 5ASimilar to the finger posture detection device 100c shown, the difference is that the finger posture detection device 100d in this embodiment omits the sheet-like extension 160. The first sensor 130d is, for example, a Hall sensor, disposed on the side of the second wearable part 120 near the grip part 52. That is, when the user wears the finger posture detection device 100d of this embodiment, the first sensor 130d corresponds to the inner side of each finger.

[0078] In this embodiment, the sensing element 54 is used as the sensing reference position of the first sensor 130d. When the finger is bent or straightened, the first sensor 130d obtains the corresponding sensing value based on the distance between itself and the sensing element 54, which includes a magnetic element. At the same time, based on the bending motion relationship of each finger joint, the bending degree of the first phalanx K1 and the second phalanx K2 is derived from the bending degree of the third phalanx K3, and finally the posture of each finger is obtained.

[0079] In summary, when the finger posture detection device of the present invention is worn on the user's hand, the position of the first sensor corresponds to the third phalanx of the user's fingers. Therefore, the first sensor can detect the posture of the third phalanx of each finger and deduce the posture of each finger through estimation of finger joint movement. By integrating the finger posture detection device of the present invention with the control handle, the degree of bending and posture of each finger can be detected. Accordingly, users can perform more complex input functions in virtual reality, thereby bringing out more diverse virtual reality interactions and allowing users to play games in a more natural way.

[0080] The correction method of the present invention can improve the cumulative error of the inertial measurement unit caused by long-term use. The user only needs to straighten four fingers at the same time and maintain it for a short period of time, and the computer can update the virtual finger posture with the real finger posture, so as to achieve the effect of immediate correction.

[0081] Furthermore, the correction method of this invention continuously updates the upper and lower boundaries of the actual bending range of the fingers relative to the palm using a computer based on real-time detection results, thereby correcting the virtual finger posture. This correction method can establish different detection ranges for different hand shapes or different gripping methods of different users, thus more accurately mapping the finger posture in the virtual reality.

[0082] Although the present invention has been disclosed in conjunction with the above embodiments, it is not intended to limit the present invention. Anyone skilled in the art can make some modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be defined by the appended claims.

Claims

1. A correction method applicable to a virtual reality system, the virtual reality system mapping a user's real hand posture to a virtual hand posture, the virtual hand posture including multiple virtual finger postures and a virtual palm posture, the virtual reality system including a control handle for being held by the user's hand, a finger posture detection device detachably surrounding the control handle and worn on the hand, and a computer, the finger posture detection device including multiple first sensors and second sensors, the multiple first sensors being located on multiple fingers of the hand, the second sensors being located on the palm of the hand, the multiple first sensors including an inertial measurement unit and a component sensing unit, the second sensor including an inertial measurement unit and a component sensing unit, the correction method including: The computer multiplies the real-time detection results of each of the first sensors by a first rotation matrix and a second rotation matrix to calculate the posture of each virtual finger. The first rotation matrix is ​​a transformation matrix between the coordinate systems of the first and second sensors, and the second rotation matrix is ​​a transformation matrix between the coordinate systems of the second sensors and the coordinate system of the control handle. The computer directly uses the real-time detection results of the control handle as the virtual hand gesture.

2. The correction method as described in claim 1, further comprising: When the difference between the instantaneous reading of the accelerometer of each of the first sensors and the instantaneous reading of the accelerometer of the second sensor is less than a preset value, and the difference between the instantaneous readings of the sensing units of the adjacent plurality of first sensors is less than another preset value, the computer corrects the virtual finger posture of the virtual hand posture to make the plurality of fingers straight and together relative to the palm.

Images