WO2011011898A1

WO2011011898A1 - Input system, and method

Info

Publication number: WO2011011898A1
Application number: PCT/CH2010/000185
Authority: WO
Inventors: David Christian Stalder; Nicolas Baumgartner; Fabian Dominik FÜRST
Original assignee: Quasmo Ag
Priority date: 2009-07-28
Filing date: 2010-07-22
Publication date: 2011-02-03

Abstract

According to an aspect of the invention, the input system comprises an input device with an accelerometer, and further comprises a position sensor for determining an input device position, and an evaluation unit, the evaluation unit being programmed to calculate, from position data provided by the position sensor, position acceleration data and to use the position acceleration data to correct accelerometer data provided by the accelerometer.

Description

INPUT SYSTEM, AND METHOD

FIELD OF THE INVENTION

The invention relates to human/computer interfaces for personal computers (PC) or consoles or set-top boxes or robots or remote control (RC) vehicles in the field of model building, such as cars and cycles, aircraft and boats. The input signals are used to control video-games or to navigate through three dimensional menus, landscapes, or virtual worlds or to navigate or control a robot or a RC vehicle in the real world . More particularly, this invention relates to an input system generating control signals in accordance with the position and the orientation of an input device operated by a user.

BACKGROUND OF THE INVENTION

For the past 30 years, video-games have mainly been played either by gamepads, 'joysticks' or by mouse and keyboard. The computer-mouse is an input device that allows very fast and very accurate control of either PC-applications or games. The problem of gaming with a mouse is that it requires a workspace, a table or another flat and hard surface to place and navigate the mouse on. The keyboard on the other hand was designed for entering characters and not as action buttons as they are used in gaming. Robots, ballbots and RC vehicles have equal conditions. In general, robots already play an important role in the industry in the manufacturing. In addition, more and more robots are used to try to get the benefits from the industry in normal life, for instance in service sector. For example at the museum, they can provide a visitor with the needed information or accompany her/him to the next exit. A s-called 'ballbot' has an additional characteristic, which makes it more difficult to control or navigate. A ballbot is a self balancing robot and moves on a single ball instead on wheels. A ballbot can move spontaneously in any direction and can rotate on a single position. According to how a ballbot moves, the ballbot itself has no point of reference for the user to know, which direction is ahead. Today the used control devices are peripherals of a personal computer, such as mouse/keyboard combination, joysticks and gamepads. Those devices where actually made for another purpose and are able to handle two axes, which means they work well in two-dimensional space or virtual worlds. For controlling or navigating a robot, the user needs an intuitive and easy to use control device, especially in the environment of humans.

Nowadays most of the users play games in the living room, in front of the TV-sets, using a game-console. Because keyboards and mice are not suitable for the living room, most of the players use so called gamepads to control the games. The gamepads have the advantage that they can be used for a wide variety of game- genres. The disadvantage is that basically all controls are handled by the two thumbs. This principle of control was adequate for old games with 2D graphics and where only simple interactions were required. Nowadays, where most games have 3D graphics, and where very complex interactions with the virtual world are required, gamepads reach the limits of the usability. The gamepads do not offer enough input axis, they are not intuitive for most of the games and they are not precise enough for accurate controlling. Due to this reason, some companies have started to develop new input devices for playing video games. First to mention is the Japanese company Nintendo, which launched in 2006 the console "Wii". The main feature of the console is the new input device, which captures the movements of the device operated by the player. This idea of motion capturing of an input device has also been taken over by several other companies. The advantage of such an input device is that more intuitive control is possible. No more only the thumbs are used to play games. Body gestures and movements are bringing more fun into the gaming. The human/computer interface is getting closer to the human side, because the input device starts interpreting natural body movements and converts them into computer inputs.

One broadly used technology to detect the tilting angle of a motion based input device is acceleration sensors or accelerometers. When a two dimensional accelerometer is used, the tilting angles (pitch- and roll-angles) can be detected because of the earth gravitational field. One problem of the accelerometer technology is that they do not only detect the earth gravitational field, but also dynamic acceleration, which occurs when the input device is moved (accelerated/decelerated) by the user. The measured dynamic acceleration disturbs the orientation measurement (pitch-, roll-angles) of the device and therefore also the control of the game or the 3D-navigation. For example, a dynamic acceleration of half the gravitational acceleration g is quite likely to occur. That may correspond to an error of the measurement of 30 degrees. Unfortunately, it is technically very hard to distinguish between earth gravitation and dynamic acceleration.

In the patent application publication US2007/225938 by Nintendo Co, a method to address this drawback is described. The described solution is to observe the magnitude of the acceleration vector and test if it is in a predetermined range around the gravitational constant. If not, the inclination calculation process is skipped. Further attempts to address the problem include using a plurality of gyroscopes for measuring the angular velocities instead of using accelerometers. Gyroscopes, however, feature the disadvantage that the angular velocity has to be integrated, and possible inaccurateness add up over time. Also, gyroscopes are more expensive than accelerometers.

SUMMARY OF THE INVENTION

It is an object of the present invention to provide an input device that overcomes the disadvantages of orientation measurements of the state of the art and especially to improve accelerometer based orientation measurements.

It is a further object to provide an angle and/or angular velocity and/or orientation determining input device.

It is a further object of the invention to provide a method of exactly determining an orientation of an input device.

According to an aspect of the invention, the input system comprises an input device with an accelerometer, and further comprises a position sensor for determining an input device position, and an evaluation unit, the evaluation unit being programmed to calculate, from position data provided by the position sensor, position acceleration data and to use the position acceleration data to correct accelerometer data provided by the accelerometer. In other words, according to the aspect of the invention, dynamic accelerations generated by movements of the handheld device may be calculated from time- dependent position data, and this may be used to compensate unwanted components of the accelerometer signal that are caused by the dynamic accelerations, resulting in corrected accelerometer data representing components of the acceleration that are due to the gravitational acceleration and that allow an orientation determination of the roll and pitch angles.

It has been found that by the relatively straightforward-to-implement measures taken by the invention, the sensitivity of the roll and pitch angle measurements to dynamic accelerations may largely be removed. Moreover, the approach according to the invention may, in contrast to approaches according to the prior art, be implemented using relatively few or almost no additional hardware, as for may applications the input system will have a positional sensor for the input device anyway. For these reasons, the input system according to the invention is very cost effective compared to prior art input devices.

Preferably, the accelerometer of the input device is at least a 2-axis accelerometer sensor, i.e. it comprises sensors for measuring the acceleration with respect to at least two axes (namely, the x and y horizontal axes).

In accordance with an especially preferred group of embodiments, the input device in addition to the accelerometer(s) comprises a rate gyroscope and/or a 3-axis geomagnetic field sensor. In addition or as an alternative, the accelerometer may be a three axis accelerometer. This, among other advantages, allows to detect upside-down orientations of the input device.

A rate gyroscope and/or a 3-axis geomagnetic field sensor with a three axis accelerometer allows to also determine the orientation of the input device along the vertical axis perpendicular to the earth's gravity field (yaw-angle). With the pitch-, roll- and yaw-angle, the orientation of the input device in the 3D-space is known.

The yaw-angle can preferably also be used as an additional input for the personal computer, console, set top box, robot or RC vehicle.

In an embodiment of the invention, calculating the dynamic acceleration includes calculating the second derivative of the input device position as a function of time.

The correction may include, according to a first option, a vectorial subtraction of the dynamic acceleration (or its projection onto the horizontal x-y-plane) from the initially determined accelerometer data. This is possible and provides good results in case the orientation of the position sensor with respect to the input device is known, so that the horizontal axes of the reference systems of the input device and the position sensor are well-defined with respect to each other, and if the pitch-, roll- and yaw-angles are comparably small. For many applications, this will always be the case, so that the first option provides a sufficient approximation.

For the relative orientation of the horizontal axes of the input device to be known, the user may be asked to place a control device containing the position sensor in a well- defined orientation. In addition or as an alternative, the input system may comprise means for determining the 3D-relative orientation of the input device and position sensor reference systems. Such means may include calibration means that use initial movements and the comparison of accordingly generated signals to calculate the orientation. In addition or as an alternative, these means may include means for determining the full 3D-orientation of the input device, together with instructions to place the control device in a well-defined orientation. In addition or as yet another alternative, these means may include means for determining the full relative position and orientation information of the input device and of the control device; this may include means for measuring the orientation of the control device as well; as will be described in more detail hereinafter.

If the pitch-, roll- and yaw-angles become too big, the axis of the dynamic acceleration, measured with the position sensor, deviates from the accelerometer axis.

If all three orientation angles (pitch, roll and yaw) are known, it is in addition possible to perform a coordinate transformation, where the coordinate system of the position sensor, is transformed into the coordinate system of the input device. According to a second option, therefore, the correction may include a coordinate system (reference system) transformation prior to a subtraction, so that at any given moment the vectorial subtraction provides an exact (with the possible exception of measurement inaccuracies, which are usually small compared to the signals) result. Due to this, the compensated tilt-data is always accurate, independent on how the user holds the input device

Also, if a three-dimensional position sensor is used, dynamic acceleration along the vertical axis (z axis) can also be detected, and therefore it can be compensated. If the position of the vertical axis is also considered as an input, one has a full six degrees of freedom input device, with an accurate compensation of all dynamic accelerations.

Now turning to the position sensor, the position sensor may be any suitable position sensing device or set-up, including a position sensor as such known from the prior art. There are many different ways to measure the position of a specific device, for example by a triangulation method. In this case, several emitters of electromagnetic or acoustic waves are placed around the input device. By analyzing the waves that are back-reflected from the input device, the distance from the emitter to the input device can be measured. If the exact position of all emitters is known, the position of the input device can be determined.

Another way to measure the position of an input device is by an electromagnetic tracking system, described in the patent application publication US 2007/0299623. The system determines the position of one three-axis coil array with respect to a second three-axis coil array using the near- field properties of the coils.

Another popular way to measure a position of a device is by optical tracking. In this case, a camera films the input device. The resulting digital image data is treated by signal processing algorithms, which allow the identification of the input device in the image, and therefore the input device can be tracked and the position determined. The disadvantage of this method is that it is very difficult to measure the distance from the camera to the input device. Recently there appeared 3D cameras, which not only capture a two-dimensional image, but also measure the distance from the camera to the object for every pixel. This technology enables three-dimensional images capturing, which leads to three dimensional position tracking. In an especially preferred embodiment of the invention, the method of measuring the position of an input device as described in the patent application PCT/CH2009/000084, incorporated herein by reference in its entirety, is used. The position sensor sends out infrared rays, using an infrared LED. The rays are reflected back from the input device. A lens is forming an image of the input device, but instead of using a CCD or CMOS camera, a so called position sensitive device is used. This device does not capture an image. From its four analogue outputs, the center of the highest light intensity can be derived, and therefore also the position of the input device. By measuring the intensity of the reflected light, the distance from the position sensor to the input device can be calculated. Compared to the method of using a camera to capture an image, the method with the position sensitive device can only measure the position of one input device at a time. On the other hand, this technology can measure the position in three dimensions, it is cheap and simple and also very precise.

The control device with the position sensor detects the position of the input device in a fixed reference system which depends on the orientation of the position sensor. If one considers that the position sensor is positioned on the floor, the only degree of freedom for the position sensor is a rotation around the vertical axis. In one embodiment, the user has to orientate the position sensor in the same way as he defines the movements of the input device in the three directions. If, in another embodiment, a geomagnetic field sensor is integrated into the position sensor, the user does not have to care how he orientates the control device on the floor, because the orientation of the position sensor and the input device (in the case that the input device also incorporates a geomagnetic field sensor) are both known and the coordinate transformation can be performed accordingly. If the user mounts the position sensor, for example, above the video screen, thus relatively high up, a built in (at least) one dimensional, acceleration sensor can be used to correct a non favorable pitch angle of the position sensor. Also in this case, the orientation of the position sensor and the input device are both known and the coordinate transformation can be performed accordingly.

The evaluation unit used to carry out the mentioned calculations, including, if necessary, the coordinate transformation, may be located anywhere in the input system (including the input device and/or the control device), or it may, though it forms part of the input system, be located in an external hardware component, for example in the computer to which the control signal is input, a settop box, a game console, etc. Also distributed arrangements of the evaluation unit are possible, for example with a first calculation step or pre-evaluation step being carried out in the control device, and further evaluation steps being carried out in the computer, settop box and/or console, robot or RC vehicle

A method of generating a control signal for a video game and/or a visual interface according to an other aspect of the invention the method comprises the steps of measuring an acceleration of an input device held by a user, to thereby obtain a measured acceleration, of determining a position of the input device, of calculating, from the determined position, a position acceleration of the input device, of correcting the measured acceleration by subtracting the position acceleration from the measured acceleration, to thereby obtain a corrected acceleration, and of determining an input device orientation from the corrected acceleration.

BRIEF DESCRIPTION OF THE DRAWINGS

In the following, preferred embodiments of the invention are described referring to drawings. The drawings are all schematical. Same reference numerals refer to same or analogous elements. In the drawings: - Fig. 1 shows a system in which the input device operates;

- Fig. 2 depicts a user holding an input device;

- Fig. 3 represents tilt data before and after compensation;

- Fig. 4 shows a visualization of the problem of the dynamical acceleration;

- Fig. 5 shows a preferred embodiment of a compensation procedure in a block diagram;

- Fig. 6 depicts a representation of a possible exterior view of the input device;

- Fig. 7 shows a schematic representation of an embodiment of the input device;

- Fig. 8 shows a schematic representation of an embodiment of a control device;

- Figs. 9-11 represent screens showing applications of embodiments of the invention; and

- Fig. 12 shows an example how a so called ballbot is controlled by the input device DESCRIPTION OF THE PREFERRED EMBODIMENTS

Fig. 1 shows the whole system in which the input device operates, including the input system. It comprises the input device himself 1, which is preferably hold in both hands by the user, and preferably has the shape of a sphere. The input device detects preferably three orientation angles, button press events and other inputs. These inputs are sent by wire or preferably by a wireless connection 30, preferably a radio communication, to the control device 2, which preferably also incorporates a position sensor and preferably the receiver for the communication. The input device together with the control device forms an input system.

The position sensor of the control device 2 is for example a position sensor as described in PCT/CH2009/000084 incorporated herein by reference.

The control device 2 is connected to a video game console or a personal computer or a set top box 3 through a data-bus 6, preferably incorporated into the control device. The communication protocol can be for example an USB, a serial, a Bluetooth or another communication protocol. The data is sent by wire or a wireless connection. A screen 4 is connected with the personal computer or the console or the set top box to visualize the application in which the control device is used. Also loudspeakers 5 are connected to a personal computer or a console or a set top box.

Fig. 2 shows a user, holding the input device 1 in his hands. The orientation-angles pitch 18, roll 19 and yaw 20 are shown in the figure with respect to a Cartesian coordinate system 22 related to the input device. An external control device with an incorporated position sensor 23, with a fixed Cartesian coordinate system 21, related to it, is also shown. A basic version of the input device can detect the two orientation angles, pitch and roll, with an, at least, 2-axis accelerometer sensor. The acceleration sensors measure the components of the earth's gravity field with a two dimensional acceleration sensor in the two dedicated directions. It is possible to determine the declination of the input device about the two dedicated axes. Pitch and roll angles can for example be calculated with formula (1) and (2).

pitch[t]

rollit] « aa"¹ (=≤^3] (2)

Where acc x and acc_y are the accelerometer output data along x and y axis measured by the acceleration sensor in the input device, and G the gravitational acceleration.

For an accurate determination of this two orientation angles, pitch and roll, one has to measure only the accelerations due to the earth's gravity field, and not accelerations due to dynamic movement of the input device.

In addition, an at least, two dimensional control device for detecting the position of the input device in the at least two horizontal directions is used.

The position data is two times differentiated with respect to time. The result corresponds to the dynamic acceleration of the input device. The result can be used to subtract it from the data of the accelerometers, in order to compensate the tilt data of the accelerometers from the dynamic acceleration of the input device.

Fig. 3 relates to a basic embodiment including a two-axes accelerometer. Fig. 3 shows the raw tilt-data 27 measured in the input device 1 with the accelerometer, which also comprises unwanted dynamic acceleration information. To measure the dynamic acceleration in the control device 2, the position data, from the position sensor in the control device, is two times differentiated 24 with respect to time. The dynamic acceleration is subtracted from the raw tilt-data, in order to obtain the compensated tilt-data 26.

Fig. 4 visualizes the problem with the dynamical acceleration. The measured acceleration vector ace 27 is composed of the Gravitational acceleration vector G 26 and the dynamical acceleration dynjacc 24. The effective roll angle α 28 is degraded by the angle β 29, if no compensation is effectuated.

A simple compensation is to subtract the dynamical acceleration calculated by the control device without a coordinate transformation. In this case input device and control device with position sensor have to be aligned and the effectuated rotations with the control device have to be small.

pttchlt] = sin^"1 (— ^") ₍₃₎ roH[_t] = w-» (=^≤±22!>£3) ₍₄₎ In the case of a two-dimensional position sensor, the dynamical acceleration from the input device in the vertical direction, are not compensated. The vertical dynamical acceleration in the same direction as the earth's gravity field is the most insignificant component for the compensation.

With respect to Fig. 5, an embodiment is described that further includes a rate gyroscope or 3-axis geomagnetic field sensor of the input device, and wherein the accelerometer is a three axis accelerometer.

If all three orientation angles (pitch, roll and yaw) are known, it is possible to perform a coordinate transformation, where the coordinate system of the position sensor 21, is transformed into the coordinate system of the input device 22. Therefore, the dynamic acceleration measured in the control device with the position sensor 24, corresponds at any given moment to the dynamic acceleration component measured with the accelerometers in the input device. Due to this, the compensated tilt-data 26 is always accurate, independent on how the user is holding the input device.

In addition, by using a three dimensional position sensor, dynamic acceleration along the vertical axis (z axis) can also be detected, and therefore it can be compensated. If the position of the vertical axis is also considered as an input, one has a full six degrees of freedom input device, with an accurate compensation of all dynamic accelerations.

Fig. 5 shows the preferred compensation procedure in a block diagram. In a first step, the input device is initialized while being held in a steady position to ensure that the initial orientation measurement is correct. In this initialization procedure 31 , an uncompensated but correct calculation of the orientation angles roll, pitch and yaw is effectuated. The acceleration-data 27 of the input device 1 is measured with the dual axis acceleration sensors. In a further step, the position of the input device is measured with the external position sensor 23, preferably incorporated into the control device 2. The position data is two times differentiated with respect to time, in order to obtain the dynamic acceleration 24. The dynamic acceleration is calculated in a fixed reference system of the steady position sensor. So, one has to express the dynamic acceleration in the moving reference of the input device. This is done with a coordinate transformation 25 by using the last calculated and yet compensated roll- and pitch-angles, plus the yaw angle. The yaw-angle is measured by a yaw-angle sensor 32 (gyroscope or geomagnetic field sensor) that is preferably incorporated into the input device. In a further step, the coordinate transformed dynamic acceleration 25 is subtracted from the accelerations 27 measured in the input device, in order to obtain the compensated orientation angles 26 pitch, roll and yaw. These angles are, together with the position data 23 transferred through a data-bus 6 to a PC, console or set-top box 3, where they are used as input signals, to control a certain application.

For defining the orientation of the input device, one can for example first rotate positively about the fixed x-axis (pitch) then positively about the fixed y-axis (roll) and at last positively about the fixed z-axis (yaw). This convention for the three orientation angles roll, pitch and yaw is an appropriate variant. This convention is used for the following formulas. For the sign convention see Fig 2. With this definition, the matrix in formula (5) can be used to calculate the coordinate transformation, in order to transform the double differentiated vector of the position data into the reference system of the input device. In this case the formula for the compensation is mentioned below. / cosrcosy — sinycosr ήnr \

G = ace— I sinpcosy— sin p sin r cosy cosp cosy +sinp sin rsin y sinpcosr ] -pas

\— sin p sin y— sin r cosp cosy — sinpcosy + sin rsin y cosp cosrcosp/

(5)

In this formula G is the acceleration vector due to the earth's gravity field, ace is the acceleration vector measured in the input device by the acceleration sensors and pos

5 is the second derivative of the position data measured by the external position sensor in the control device with respect to the time, expressed in the reference system of the control device, which is transformed with the aid of the rotation matrix in the reference system of the input device. In the rotation matrix we us r, p and y as abbreviations for the compensated roll, pitch and yaw angles calculated one time step 10 before.

The pitch- and roll-angles are then obtained by the following formulas, in respect with the above mentioned convention, where G is the absolute value of G.

pitcfoξέ] = sin

(7)

, ¹⁵ _s roll [ ^Lt] ^J = sin^"1 f VG-cod ftpi^*tcftJ*] Λ)/ , ⁽8„) The yaw angle can preferably be measured with the aid of a three-axis geomagnetic field sensor, incorporated into the input device. The geomagnetic field vector is projected into the horizontal plane, perpendicular to the gravity vector. For doing that, the magnetic field components (9), which are measured in the reference system of the input device, have to be transformed. The yet calculated pitch- and roll-angles can be used for this transformation. The yaw-angle can then be calculated by the transformed components b_y' and b_x' of the geomagnetic field with the arc tangent (12).

→ ( ^■ cosr 0 sJnr \

BB'' == 1 I s siϊnn psinr cosp — sin p cosr J - B (iθ)

\— s ήinrcosp sinp cosr cosp J

In the formulas, mentioned above, B is the geomagnetic field vector measured in the reference system of the input device (9). In the formula (10) B^r is the tilt compensated geomagnetic field vector. Its components are shown in formula (1 1). Hereinafter, preferred features of embodiments of the invention are described in some more detail.

One preferable shape of the input device is the shape of a sphere or a shape reminding of a sphere. The preferable way to hold it is with two hands. In this case, the user is able to handle the device precisely and in a relaxed position. It is also preferable that the input device features buttons to perform further input commands like for example an accept-button, a fire-button or an jump-button. Preferably these buttons are arranged on the surface of the input device in a way, that if the input device is held by the user in a relaxed way, every finger or a plurality of fingers each can manipulate one button. This way all necessary buttons can be manipulated without any need for the user to move his hands or fingers, when he is holding the input device. These buttons are preferably pressure sensitive, or they are built as triggers to distinguish whether the user is performing a low pressure interaction or a high pressure interaction, and this information is then used as a further input. For example, a low pressure button interaction could be interpreted as a little jump action of the virtual world avatar, and a high pressure button interaction as a high jump action of the avatar. The buttons preferably incorporate a feedback device that gives the user a feedback of how much pressure he applied at the action button. That feedback could be a force that is acting in the opposite direction of the button/trigger movement, or a locally applied vibration of the button. The more pressure is applied the stronger the vibration.

The input device preferably incorporates a so called force-feedback device. This device is controlled by the game or the application. The force-feedback device gives information to the user by vibrations. For example, if a gun is fired, the force- feedback device vibrates for a short time to simulate the repulsion of the gun. The input device preferably incorporates a loudspeaker. The loudspeaker emits sounds and noises that are specific to the avatar or the user. For example the sound of pushing the trigger of a gun or the sound of a voice giving the user hints how to perform better in the game or application.

The input device preferably incorporates a microphone which allows the user the communication with other players in an online game. It could also be imagined that a voice recognition unit is interpreting the user's voice as input commands.

The input device preferably incorporates a display with a touch or multitouch sensitive surface. The display shows additional information related to the running software application. Due to the touch sensitive surface, the user is able to select different options related to the running software application. For example, he can choose in which location data is stored, or which user profile should be loaded.

In Fig. 6, a representation of a possible exterior view of the input device 1 is shown. The action buttons/ triggers 10 are shown. They are arranged on the surface of the input device in a way, that every finger of the user's hands is able to press one specific button/ trigger, without that the user has to move his hands. A display with a touch-sensitive surface 1 1 is mounted into the input device, in a way that it faces the user when he is holding the input device. The surface of the input has some holes, beyond which a microphone 12 and a loudspeaker 13 are mounted.

Fig. 7 shows a schematic representation of an embodiment of the input device 1. The subsystems, which are mapped on the circle (representing the input device), are the peripherals of the input device. A microphone 12, a loudspeaker 13, buttons/triggers 10 and a display with touch sensitive surface 11 are shown. Inside the input device, the different subsystems, like a three-axis accelerometer 27, a geomagnetic field sensor 32, a force-feedback system 17 and radio-frequency communication system 30, are represented. In the center of the figure, a microcontroller 33 is shown. The microcontroller manages the interaction with all sub-systems of the input device. The arrows represent, in which way the microcontroller interacts with the sub-systems.

Fig. 8 shows a schematic representation of an embodiment of the control device 2. A three-axis accelerometer 35 and a geomagnetic field sensor 35 are shown. These subsystems are used to detect the orientation of the control device, in order to effectuate a correct coordinate transformation for best compensating results. The position sensor 23 detects the position of the input device and communicates it to the microcontroller 34, which controls the interaction with all sub-systems incorporated into the control device. Further, the microcontroller does all the required calculations (coordinate transformation 25, compensation 26, etc.). The radio frequency communication unit 30 is responsible for the communication with the input device 1 , and the data-bus 6 is responsible for the communication with the PC/console/set-top box 3.

Examples of applications are described referring to Figs. 9-11.

Fig. 9 represents a screen showing an application of a video-game. The user can move a virtual avatar 7 through a three dimensional world. The movements of the avatar are controlled by the user by translating and rotating the input device 1. For example, if the avatar should walk straight or backwards, the input device 1 is moved further away or closer to the user, respectively. To turn the avatar around, the user performs a yaw-movement with the input device 1. To change the viewing angle of the camera, the user performs a pitch-movement with the input device 1. For sidestepping, the user moves the input device to the left or to the right side. If the avatar should crouch or jump, the user lowers respectively holds down or up the input device. To perform further actions with the avatar, like for example performing a strike action with a sword 8, or a defending action with a shield 9, the action buttons/triggers 10 on the surface of the input device can be used.

Fig. 10 represents a screen, showing an application of a video-game. The user can pilot a virtual airplane and fly pitch-, roll- and yaw-movements by performing pitch-, roll- and yaw-rotations of the input device 1.

Fig. 11 represents a screen showing an application of a three-dimensional visual interface. The visual interface could be, for example, a so called media center. A media center is an application that allows the user rapidly access all his digital media content, like movies, pictures, music-files and so on. The content 16 could be organized in different windows 15, arranged in three dimensions. With specific movements of the input device 1 , the user can very quickly navigate through all his media content, and therefore quickly find the desired media file.

Fig. 12 represents a ballbot. The user can move the ballbot 36 in the real world. The movements of the ballbot are controlled by the user by translating and rotating the input device 1. For example, if the ballbot 36 should walk straight or backwards, the input device 1 is moved further away or closer to the user, respectively. To turn the ballbot 2 around, the user performs a yaw-movement with the input device 1. For sidestepping, the user moves the input device 1 to the left or to the right side. If the ballbot 2 should raise or decrease itself, the user lowers respectively holds down or up the input device 1. To perform further actions with the ballbot, like for example photograph an object, the action buttons/triggers 10 on the surface of the input device 1 can be used. Various other embodiments can be envisaged without departing from the scope and spirit of the invention.

Claims

WHAT IS CLAIMED IS:

1. An input system for generating a control signal for a video game and/or a visual interface and/or a robot and/or a remote control vehicle, the input system comprising an input device to be held by a user, the input device comprising at least one accelerometer, the input system further comprising a position sensor for determining an input device position, and an evaluation unit, the evaluation unit being programmed to calculate, from position data provided by the position sensor, position acceleration data, and to use the position acceleration data to correct accelerometer data provided by the accelerometer to yield a corrected input device orientation.

2. The input system according to claim 1, wherein the at least one accelerometer is/are capable of determining the acceleration with respect to at least two distinct directions.

3. The input system according to claim 2, wherein the at least one accelerometer is/are capable of determining the acceleration with respect to all spatial directions.

4. The input system according to any one of the previous claims, wherein the input device further comprises a rate gyroscope and/or a 3 -axis geomagnetic field sensor.

5. The input system according to any one of the previous claims, wherein the evaluation unit is programmed to calculate the second derivative of the input device position with respect to the time in order to calculate the position acceleration data.

6. The input system according to any one of the previous claims, wherein the evaluation unit is programmed to perform a coordinate transformation of an input device coordinate system into a position sensor coordinate system, or vice versa, prior to correcting the accelerometer data.

7. The input system according to any one of the previous claims, wherein the position sensor is arranged in a control device, and wherein the control device comprises means for determining an orientation of the control device with respect to at least one axis.

8. The input system according to any one of the previous claims, wherein the position sensor comprises an electromagnetic radiation source and a position sensitive device for capturing radiation emitted by the electromagnetic radiation sensor and reflected by the input device.

9. The input system according to any one of the previous claims, the input device further comprising at least one of the group comprising at least one input button, at least one pressure selective input button, at least one force feedback device, at least one loudspeaker, at least one microphone, at least one display and at least one display with a touch sensitive surface.

10. A method of generating a control signal for a video game and/or a visual interface and/or a robot and/or a remote control vehicle, the method comprising the steps of measuring an acceleration of an input device held by a user, to thereby obtain a measured acceleration, of determining a position of the input device, of calculating, from the determined position, a position acceleration of the input device, of correcting the measured acceleration by subtracting the position acceleration from the measured acceleration, to thereby obtain a corrected acceleration, and of determining an input device orientation from the corrected acceleration.

11. The method according to claim 10, comprising the additional step of performing a coordinate transformation prior to the step of subtracting the position acceleration.

12. The input system according to any one of the previous claims, the input device has the shape of a sphere or a shape approximating a sphere.