`European Patent Office
`Office europeen des brevets
`
`© Publication number : 0 4 8 4 1 60 A 2
`
`EUROPEAN PATENT A P P L I C A T I O N
`
`19
`
`12
`
`© Application number: 91310077.2
`
`(£) Int. CI.5: G06K 11/08
`
`@ Date of filing : 31.10.91
`
`© Priority: 02.11.90 US 608439
`(43) Date of publication of application
`06.05.92 Bulletin 92/19
`
`@ Designated Contracting States :
`DE FR GB
`
`© Applicant : XEROX CORPORATION
`Xerox Square
`Rochester New York 14644 (US)
`
`@ Inventor : Elrod, Scott A.
`262 Hawthorne Avenue
`Palo Alto, CA 94301 (US)
`Inventor : Tang, John
`865 Robb Road
`Palo Alto, CA 94306 (US)
`Inventor : Minneman, Scott L.
`1550 Noe Street
`San Francisco, CA 94131 (US)
`Inventor : Jackson, Warren B.
`160 Castenada Avenue
`San Francisco, CA 94116-1407 (US)
`© Representative : Hill, Cecilia Ann et al
`Rank Xerox Patent Department Albion House,
`55 New Oxford Street
`London WC1A 1BS (GB)
`
`© Position and function input system for a large area display.
`
`© An input device for simultaneously entering
`position and function information into an elec-
`tronic system having a viewing surface (20)
`upon which is displayed information generated
`by the electronic system. A light pen (22) is
`provided for projecting a light spot onto the
`viewing surface for indicating the position of a
`location marker thereon and for indicating a
`function to be performed. The optical output of
`the light pen is modified to represent the selec-
`ted function. A sensor (28) receives the light
`in conjunction with discrimination
`spot and,
`electronics, determines the location of the cen-
`troid of the light spot relative to the viewing
`surface, and determines the function to be per-
`formed.
`
`si
`\
`\
`
`o
`(
`
`is
`\
`V
`\ \
`\ \
`\ \
`\
`1
`
`CM
`<
`O
`CO
`
`00
`
`LU
`
`Jouve, 18, rue Saint-Denis, 75001 PARIS
`
`IPR2015-01347
`PA 1008 Page 001
`
`
`
`EP 0 484 160 A2
`
`5
`
`w
`
`15
`
`20
`
`This invention relates to a wireless optical input device for simultaneously, in a single beam, entering posi-
`tion and function information into an electronic system. Plural, uniquely encoded input devices are particularly
`suitable for use in a collaborative system, wherein multiple persons may work together for supplying input infor-
`mation to a single large area display.
`Computer systems generally incorporate a display unit for providing a visual indication to the user of selec-
`ted data. A specific location marker, such as a pointer, may be moved by the user to any desired point on the
`display in order to locate a cursor for the entry of keystroke characters, to trace the locus of points as in drawing
`alphanumeric characters or other patterns, to invoke and manipulate a functional command such as paint or
`erase, to open a menu, to invoke a displayed command, or other interface functions. In each case the location
`of the pointer must be known and in many applications the desired control function should be known as well.
`Pointer positioning, as a computer input device, has been commonly effected in a variety of different ways.
`For example, by designated keys on a keyboard, by a freely movable "mouse" having one or more function
`selection buttons thereon, by a "joystick", and by means of a stylus upon a graphics input tablet. Each has its
`own unique advantages. Keyboard input allows the user to designate a location without removing hands from
`the keyboard, a mouse is easily and rapidly movable over a pad in correspondence to the display area and its
`function selection buttons allow various common subroutines to be invoked, the joystick is also a rapid posi-
`tioning device, and the stylus enables freehand input.
`The known light pen is another computer input device which gives the user the direct interactive "feel" of
`drawing on the display surface. It is usually in the nature of a receiver (not a transmitter), for use with a rastered
`video display screen, and receives timing information from the raster scan. Therefore, it is hard wired to the
`computer for transmitting the received signals thereto so that the timing information may be translated into
`positional data which, in turn, is used by the computer software to control the position of the pointer on the
`screen.
`In a collaborative working environment, where several users wish to view and manipulate displayed infor-
`25 mation simultaneously, it is desirable to provide a large area display measuring several feet across (both hori-
`zontally and vertically). Each of the multiple users of would manipulate a light pen which could be used
`simultaneously and independently for controlling its related pointer on the display in order to position a cursor,
`select an item from a menu, draw upon the display screen, or perform any of a number of standard functions.
`In this way the actions of each user would be readily visible to all the members of the group who would interact
`together much as they would relative to a chalkboard. Clearly, when used in this manner the hard wired, re-
`ceiver-type light pen would be a hindrance to ease of communication.
`A direct input device for such a large area display system should desirably comprise a wireless light pen
`emitting optical radiation which could be detected behind the display screen. It should be equally usable, relative
`to the display screen, as a remote pointer, by users comfortably seated several feet from the screen, as well
`as in "writing" contact with the screen. It should be capable of pixel location accuracy and should be carefully
`designed for environmental safety so that, at normal distances of use, its optical beam would be incapable of
`focusing a light spot on the eye and causing eye damage.
`Despite the seemingly inconsistent requirements that the light beam should not cause eye injury, and yet
`be able to identify accurately pixel locations when operated several feet from the display screen, it is never-
`theless an object of this invention to provide a system/method which will enable a noninjurious beam of optical
`radiation incident upon a display screen to be resolved to the pixel level.
`It is a further object of this invention to provide an interactive display system capable of simultaneously
`receiving the optical input of plural sources of illumination, for example plural wireless light pens, wherein the
`input of each will identify its projected location, at the pixel level, and each may be used to invoke plural func-
`tions.
`The present invention provides one or more input devices for simultaneously and independently entering
`position and function information into an electronic system comprising a large area viewing surface upon which
`is displayed information generated by the electronic system. The output illumination of each input device
`uniquely identifies the source and the function to be performed and is projected as a light spot upon the display
`surface. All of the projected illumination falls upon a sensor which generates output signals representative of
`the total optical input of the light spots. These output signals pass through discrimination electronics for generat-
`ing signals representative of the locations of the centroids of each of the light spots relative to the viewing sur-
`face and signals representative of the identified functions to be performed.
`By way of example only, embodiments of this invention will be described with reference to the accompany-
`ing drawings wherein:
`Figure 1 is a schematic representation of a projection display system and a light pen digitizing system,
`Figure 2 is a side sectional view through a light pen,
`Figure 3 is a schematic representation of three light pens and their clustered function frequencies,
`2
`
`so
`
`30
`
`35
`
`40
`
`45
`
`55
`
`IPR2015-01347
`PA 1008 Page 002
`
`
`
`EP 0 484 160 A2
`
`5
`
`w
`
`15
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`so
`
`55
`
`Figure 4 is a block schematic of the light pen electronics,
`Figure 5 is a block schematic of the sensing and cursor control electronics for a single light pen,
`Figure 5a is a timing diagram associated with the sampling electronics,
`Figure 6 is a schematic representation of an alternative encoding arrangement for the light pens (one
`shown),
`Figure 7 is a schematic representation of another encoding embodiment for source and function identifi-
`cation of the light pens, and
`Figure 7a is a schematic representation of two alternative combined input signals at different phase rela-
`tions.
`Turning now to the drawings, there is illustrated in Figure 1 a large area display terminal 10 in the form of
`a rear projection system comprising a one million pixel liquid crystal light valve panel 12, controlled by a com-
`puter (not shown). The panel 12 is interposed between a high intensity projection lamp 14, such as a 650 watt
`Xenon arc lamp, focused by fresnel lens 16, and a 270mm projection lens 18. The image is magnified about
`fivefold to illuminate, at about twenty spots per inch (one inch = 2.54 cm), a slightly convex curved (as viewed)
`display screen 20 having an area of about three feet by five feet (91 .5 cm by 1 52.5 cm). The screen 20 is pref-
`erably made of plastic and is etched to diffuse light for providing sufficient non-directionality to the viewing angle
`so that viewers need not be located directly in front of it.
`One or more wireless light pen 22 (two shown) projects a beam of infrared (IR) light from a light source
`such as an LED onto the front surface of the screen 20 at a location where the user desires to indicate input,
`such as, locating a pointer. Being wireless, the pen has enhanced usability as a collaborative tool since the
`pens can be used as light spot projection devices at optimum distances between the screen surface and several
`feet from it. When the user is writing upon the screen it is preferable to maintain the light pen in contact with
`the surface being written upon. However, when the user is merely pointing or activating window or menu items
`upon the screen, it would be quite practical to project the light spot from several feet away from the screen. It
`should be noted that as a remote pen projects a larger light spot, the effective zone of accurate usage gets
`closer to the center of the screen because too much light falls off the screen. With wired pens and multiple users,
`the wires would probably get tangled in this collaborative mode of usage.
`IR projected light is used, together with an IR sensitive sensor (to be described), so as to minimize the
`amount of interference from spurious light sources. For safety reasons, the light beam projected from the light
`pen 22 should be significantly divergent, rather than being collimated. A light pen having a projection half-angle
`at half-intensity of about 25°, held about two feet from the screen, will project a usable IR light spot about two
`feet in diameter upon the three by five foot screen (one foot = 30.48 cm).
`As the IR light spot on the screen is not in the visible range, the user's feedback is solely the feedback
`generated by the electronic system and presented on the display from the information which is obtained by the
`sensor and suitable electronics. A large curvature demagnification lens 24, of about 90x magnification, directs
`the IR light spot through an IR filter 26 which blocks out spurious light and then focuses the spot upon a position
`sensing photodiode 28, such as the UDTSC25D quadrant detector, from United Detector Technology of Hawth-
`orne, CA. This device has its highest sensitivity in the IR range and is a continuous dual axis position sensor
`that provides both X and Y axis position information. It senses the centroid of a light spot and provides con-
`tinuous analog outputs as the spot traverses the active area. Suitable electronic instrumentation (to be des-
`cribed) allows the X,Y coordinates to be separated and displayed as a pointer upon the projection screen 20.
`The two foot diameter light spot noted above, when demagnified through lens 24, appears as a projection of
`about 0.3 inch in diameter on the 0.74 inch by 0.74 inch active surface of detector 28.
`A light pen 22 is more clearly shown in Figure 2. It comprises a tubular body 30 about 0.75 inch in diameter,
`comparable in size to a whiteboard dry erase marking pen. It is cordless and is provided with its own power
`source, in the form of two nickel-cadmium rechargeable batteries 32 which, connected in series, generate about
`2.4 to 2.7 volts. The optical output of the light pen emanates from four LED light sources 34a, 34b, 34c and
`34d (only two shown), such as HEMT-3301 from Hewlett-Packard of Palo Alto, CA, mounted at the front of the
`pen around a generally conical tip 36. Each of the four LEDs emits about 6mw of power, cumulatively about
`25mw, resulting in sufficient intensity to enable a high signal-to-noise ratio. At such a power level, care must
`be taken to diverge the light to prevent eye damage. To this end each LED has a cover lens, resulting in the
`aforesaid projection cone half-angle at half-intensity of about 25°. At the rear end of the pen there is a suitable
`recharging connection 38 and a disabling switch 40 which electrically disconnects the light sources from the
`batteries when the pen is seated within a recharging recess in a recharging tray (not shown). The recharging
`tray is preferably mounted directly adjacent to the projection screen 20 in order to conveniently house the light
`pens and to maintain them at maximum charge at all times.
`Three function selection buttons 42, 44 and 46 (front, middle and rear), comparable to mouse buttons, are
`conveniently located at the front of the light pen to be easily accessible to the user during manipulation thereof.
`3
`
`IPR2015-01347
`PA 1008 Page 003
`
`
`
`EP 0 484 160 A2
`
`Of course, more or fewer function selection buttons may be provided. Protruding from the conical tip 36 is a
`pin 48 sheathed in a Delrin sleeve 50 for ease of sliding movement over the display screen 20 and to limit
`scratching of the plastic screen surface. Contact of the tip with the display activates a tip contact switch 52, via
`the pin 48, which invokes the same function command as the front button 42, to which it is connected in parallel.
`When several light pens are used simultaneously, the output signal of each one must be differentiate from
`that of the others so that "marks" of one user are not confused with those of another. Additionally, since each
`pen is provided with three function selecting control buttons 42, 44 and 46, the output signal representative of
`each of these, plus a "tracking" function (i.e. light source ON with no button depressed and pen movement is
`tracked by the cursor), must be differentiate from all other signals. In order to differentiate the light pen of each
`user and the several functions invoked by each, the light sources are encoded, as by chopping, so that their
`ouputs are at different frequencies. As illustrated in Figure 3, the four frequencies, representative of button
`states, are closely clustered (about 1 % apart) and the mean frequency for each pen is sufficiently remote from
`the others so as to be accurately differentiate at sufficiently high speed. The mean frequencies for pen iden-
`tification are easily electronically distinguishable from one another by bandpass filtering, while the closely clus-
`tered, function identifying frequencies are sufficiently different from one another to be differentiate with a
`frequency-to-voltage converter circuit and subsequent comparators.
`It has been found to be desirable to modulate the light sources at rates above 1000 Hz. For example, the
`four PEN #1 frequencies could be clustered around 4480 Hz, the four PEN #2 frequencies around 5830 Hz,
`and the four PEN #3 frequencies around 7650 Hz. More specifically, the four button state frequencies could
`be as shown in the following table:
`
`PEN#1
`
`4539 Hz
`
`4500
`
`4461
`
`4422
`
`Track
`
`Middle
`
`Rear
`
`Front
`
`PEN#2
`
`PEN#3
`
`5907 Hz
`
`7752 Hz
`
`5856
`
`5805
`
`5755
`
`7 6 8 4
`
`7617
`
`7551
`
`In order to generate the several required frequencies, the LEDs of each pen are controlled by a chopping
`circuit 54 mounted upon a circuit board 56 housed within the pen, as illustrated in Figure 4. A crystal oscillator,
`causing a crystal 58 to resonate at its natural frequency, comprises the crystal, its biasing resistor 60 and biasing
`capacitors 62, and an inverter comprising a single NOR gate of a 74HC02 high speed CM05 chip 64. Based
`upon the exemplary frequencies, set forth in the above table, and the dividing circuitry (described below), the
`three custom quartz crystals (from Hi-Q of Olathe, KS) which have been selected are 2.05184 MHz, 2.67014
`MHz and 3.50370 MHz.
`The CMOS chip 64 includes four independent NOR gates. Two additional NOR gates of the chip are used
`for button identification. Three inputs R, F and M, from the rear, front and middle buttons, to two of the NOR
`gates enable an output of two bits (A and B), from pins 10 and 13, for identification of the necessary four button
`states (i.e. front, middle, rear and tracking). A dividing circuit comprising three 74HC161 high speed CMOS
`counter/divider chips 66, 68 and 70 divides down the crystal resonant frequency to the four closely clustered
`frequencies, based upon the variable A and B bits fed into the first counter/divider chip 66 and the fixed inputs.
`The 74HC series chips have been selected because they will operate at low voltage output of the light pen bat-
`teries.
`The first two counters/dividers 66 and 68 are combined in order to divide the crystal clock signal fed to each,
`based upon the A and B variable inputs from the button identification chip 64 (to pins 3 and 4 of counter 66),
`the fixed inputs 1, 1 (to pins S and 6 of counter 66) and the fixed inputs 0, 0, 0, 1 (to pins 3, 4, 5 and 6 of counter
`68). Their collective count will result in a divisor, which will be 1 13, 114, 115 or 116, depending upon the four
`possible A and B inputs of 1 1 , 01 , 1 0 and 00, which then will be fed to the last counter/divider 70 whose second
`(of four) output port (pin 13) represents a further division by 4. Thus, in PEN #1 the 2.05184 Mhz input to the
`cascaded counters would be divided by 452, 456, 460 or 464 depending upon which function is invoked. The
`resultant signals of 4539 Hz, 4500 Hz, 4461 Hz and 4422 Hz enable and disable the transistors 72 (2N2222A)
`for chopping the light output signal (i.e the modulation is from ON to OFF) from light sources 34a, 34b, 34c and
`4
`
`5
`
`w
`
`15
`
`20
`
`35
`
`40
`
`45
`
`50
`
`55
`
`IPR2015-01347
`PA 1008 Page 004
`
`
`
`EP 0 484 160 A2
`
`5
`
`20
`
`25
`
`30
`
`35
`
`w
`
`34d.
`As has been described, the multiple signals from each light pen are achieved by chopping the optical output
`of the pens to produce a family of closely clustered frequencies. It then becomes necessary to electronically
`discriminate among the various frequency components in the photodiode output. The circuit of Figure 5
`schematically illustrates one such technique wherein narrowband filters are used to separate the X and Y posi-
`tions of one light pen from the X and Y positions of another.
`Although only a single pen, operated at a single frequency is shown and described, it should be understood
`that a number of pens (e.g. three) can be used simultaneously and independently.
`The position sensing photodiode 28 includes four (two opposed pairs) electrodes 76 (X + ), 78 (X-), 80 (Y
`+ ) and 82 (Y-) each of which generates a current signal as a function of the light intensity and position of the
`centroid of the light spot projected thereon. If several pens are being used, they simultaneously project optical
`signals chopped at different frequencies. The output signals from the detector electrodes will be a complex
`superposition of square waves at those frequencies. These complex waves will be separated in the circuit des-
`cribed below wherein only representative signals for the X-axis are shown. It should be understood that the
`15 Y-axis signals are handled in a similar manner.
`X+ and X- square wave current signals are converted to voltage signals and amplified at amplifiers 84 and
`86. Initially the principal noise in the system is the detector noise so care is taken to amplify the signal to a
`usable level without introducing noise to the signal. These front end amplifiers are ultra-low noise devices, OP-
`27G from Analog Devices of Norwood, MA. Then both signals pass to standard sum and difference amplifiers
`88 and 90 for determining location. The sum of X+ and X- will always have the same phase relationship to the
`pen modulation and will be a fairly large signal, while the difference can either be in phase (on one side of the
`center of the detector) or 180° out of phase (on the opposite side of center). Next, the Xsum and Xdiff signals,
`which include fundamental and higher level harmonic frequency components (since they are comprised of
`square waves), are each passed through a switched capacitor narrow bandpass filter 92 and 94 tuned to a very
`narrow predetermined frequency range by a crystal controlled clock 95 so as to pass the cluster of frequencies
`for a specific pen. The outputs of the bandpass filters are sine waves at the first harmonic frequency. Once
`again the output signals are amplified by amplifiers 96 and 98 in order to be able more easily to extract amplitude
`and frequency information representative of position and function.
`If more than a single light pen is used, the pen identification cluster of frequencies for a given pen could
`be separated by use of a bandpass filter whose center frequency would be clock controlled for varying the signal
`which will be passed. In this manner, the Xsum and X^ (as well as the Ysum and Ydiff signals) would be scanned
`and the clustered family of frequencies would be sequentially passed. It would be better, however, to use dedi-
`cated bandpass filters, each tuned to the expected mean frequency of the clustered family, in order to keep
`the speed of the system high. There is a significant transient which occurs when the switched capacitor filter
`center frequency is altered by changing the clock frequency.
`The Xsum and Xdiff sine wave signals then pass to sample and hold circuits 100 and 102 controlled by a
`signal shunted from the Xsum sine wave. The shunted signal (C) (note Figure 5a) fires a zero crossing detector
`104 so that each time the sine wave (C) crosses zero, the output signal changes between low and high (0 to
`5 volts) as represented by signal (D). When signal (D) goes from high to low it fires a crystal controlled time
`delay circuit 106 whose output is signal (E), a negative-going pulse, approximately 5p.s in duration, which coin-
`cides with the next peak of the sine wave. The time delay is set to correspond to 1/4 cycle of the mean frequency
`of the cluster. Signal (E) controls the sampling of the sample and hold circuits 100 and 102, (AD583 from Analog
`Devices of Norwood, MA) so that at every negative going pulse a peak is sampled. Since the Xsum, X^, Ysum
`and Ydiff signals are all generated by a single light pen, it is sufficient to generate a single timing signal (E) for
`all of these signals. The output signal (F), from the sample and hold circuits, is a stairstepping DC voltage indi-
`cative of the amplitude of the Xsum and Xdiff sine wave signals (C) and (C), and representive of the light spot
`position. Final RC filters 108 and 110 remove noise from the DC signal (F).
`The DC signals (F) pass to an analog multiplexer 1 12 which scans them and sequentially passes the Xsum,
`Xdiff, Ysum and Ydiff signals through a unity gain buffer 113 to an A/D converter 114 which converts each sequen-
`tially received analog voltage signal (F) and converts it into a fourteen bit digital signal. In the simplest imple-
`mentation, the 14 bits (two bytes) for each Xsum, Xdiff, Ysum and Ydiff signal are passed by an RS232 digital
`controller 1 16 to the host computer, such as a SPARCStation-l from Sun Microsystems, of Mountain View, CA
`along with the single byte characterizing the button state. Thus, nine bytes are used for a single data point for
`a single pen. The sampling time for each data point is about 0.01 1 sec.
`The square wave signal (D) is additionally used to differentiate among the closely clustered frequencies
`to determine the invoked function. A portion of the signal is tapped off and sent to a frequency-to-voltage con-
`verter 1 18 whose output passes to four comparators 120, 122, 124 and 126, each set at a different threshold
`in order to determine the exact frequency of the signal for identifying the button state of the light pen.
`5
`
`40
`
`45
`
`50
`
`55
`
`IPR2015-01347
`PA 1008 Page 005
`
`
`
`EP 0 484 160 A2
`
`5
`
`Output from the controller 116 feeds back channel select commands to the multiplexer 112, convert com-
`mands to the A/D converter 114, and feeds the nine bytes of data point information for the single pen to the
`host computer.
`Since both sum and difference signals vary linearly with respect to the intensity of the light spot, a division
`step will yield generalized X and Y values (X and Y).
`X = ^ a n d 7 = - ^
`Xsum
`Ysum
`_
`_
`X and Y eliminate light intensity variability owing to battery power shifts, the angle at which the light pen
`is held with respect to the screen, and the distance of the light pen from the center of the screen. While X and
`10 Y have a one-to-one correspondence with the X,Y location on the screen, they are generally non-linear with
`respect to the latter. This arises from non-linearities in the imaging lens, screen curvature, non-linearities intrin-
`sic to the detection electronics, and other factors. A calibration procedure is used to convert X and Y to real X
`and Y coordinates. A regular grid of points is displayed on the screen and the X, Y value of each point is
`measured by sequentially sampling the points as a light pen is held at that grid location. Typically, for good
`calibration, about 200 sampling points on the three by five foot (91.5 cm by 152.5 cm) screen are needed. X,
`Y values for patches of the screen are fitted to X, Y coordinates using cubic splines. A subsequent linear inter-
`polation is used to generate a lookup table for the computer.
`The preferred circuitry for a system using three light pens fans out from the X and Y sum and X and Y dif-
`ference square wave signals. Each of these four signals is input to each of three dedicated narrow bandpass
`filters (one for each mean value of the cluster of pen frequencies). The four output signals from each narrow
`bandpass filter are then amplified and the three groups of signals are passed to three sample and hold circuits,
`each controlled by a separate time delay circuit (because the mean values of the frequency clusters are diffe-
`rent). The twelve output DC signals are input to the multiplexer 112, the A/D converter 114 and then to the host
`computer. Additionally, there will be three frequency-to-voltage converters and four comparators for each con-
`verier. In this manner, the complex X and Y wave forms from the position sensing photodiode 28 are separated
`into the positions and button states of each of the light pens.
`Although the above description refers to a specific frequency generating circuit and a specific discrimination
`circuit, it is possible that other suitable circuits could be used. For example, for use with multiple pens, the
`RS232 controller 1 16 may be eliminated and a microcontroller substituted therefor, whose output could be sent
`to the host computer across an RS232 link or be direct memory access (DMA).
`The illustrated embodiment relates to a projection-type computer display but it should be understood that
`the nature of the information display may take other forms as long as the light spots may be projected theret-
`hrough and collected upon a sensor. For example, the display may be a large area LCD or even a projected
`slide. In the latter case, the computer of the present system would generate the image of a pointer whose activity
`35 would be superimposed over the non-computer generated display.
`Further changes are also contemplated. Instead of a single position sensor receiving the illumination from
`all