JP3068631B2 - Coordinate input device and coordinate input method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial applications]

The present invention provides a coordinate input device that emits ultrasonic waves having different group velocities and phase velocities by a coordinate pointing device, and recognizes the indicated coordinates by measuring the propagation time reaching each of a plurality of sensors installed at various places of the propagation body. It relates to a coordinate input method.

[Prior art]

As an applied device for ultrasonic distance measurement, the coordinate indicated by emitting ultrasonic waves with different group velocities and phase velocities by using a coordinate pointing tool and measuring the propagation time difference that reaches each of multiple sensors installed in various places of the propagation body. There is a coordinate input device for recognition. FIG. 9 shows an external view of the coordinate input device. In FIG. 9, reference numeral 9 denotes a coordinate pointing device (hereinafter, referred to as a "pen") that incorporates a piezoelectric element and emits desired ultrasonic waves having different group velocities and phase velocities from the tip.
Reference numeral 0c denotes a sensor which is a receiver for receiving an ultrasonic signal emitted from the pen 9 via a propagator, 11 denotes a propagating medium of the ultrasonic signal, and a propagator of an ultrasonic wave composed of aluminum, glass, or the like; Reference numeral denotes an anti-vibration material provided for the purpose of preventing reflected waves other than the direct wave from the pen 9 from entering the sensors 10a to 10c. For example, when the pen 9 is driven as shown in FIG.
In the sensor 10, an output waveform is obtained as shown in FIG. In the ultrasonic coordinate input device having the above configuration, 2
The coordinate value of the designated point is calculated from the distance between the designated point and the plurality of sensors 10a to 10c (determined from the propagation time of the ultrasonic wave from the pen 9). Although various methods for calculating the distance γ between the sensors 10a to 10c and the designated point have been devised, basically, the distance γ is calculated from the arrival time of the ultrasonic signal emitted from the pen to the sensors 10a to 10c. In this case, for example, the detection unit shown in FIG. 12 is connected to the sensors 10a to 10c, and the reception waveform of the sensor 10 is converted to the first-stage amplifier circuit 13.
The signal is amplified by the full-wave rectifier circuit 14, full-wave rectified by the full-wave rectifier circuit 14, unnecessary frequency components are removed by the low-pass filter 15, and differentiated by the differentiator circuit 16. By defining the arrival time tg of the county, the distance γ can be calculated as γ = (vg * tg) + (offset value). When time detection is performed based on the envelope in this way,
To some extent, fluctuations are inevitably caused by the magnitude of the signal output and the filter characteristics. For this reason, when further obtaining the accuracy, it has been performed to detect the phase at an appropriate position and obtain the phase arrival time tp to obtain a finer accuracy. A detection unit (signal detection system circuit) connected to the sensor 10 that actually measures the ultrasonic signal arrival time (tg, tp) in this case.
Fig. 13 shows the configuration of Fig. 13 and Fig. 13 shows the operation timing chart of each part.
It is shown in Figure 14. In FIG. 13, reference numeral 19 denotes a preamplifier for preventing reception noise from being mixed into the sensor 10 installed in the vicinity of the sensor 10, and a detection signal (FIG. 14 "a") detected by the sensor 10. ) Are amplified by the preamplifier 19. This received signal is
It is rectified by the absolute value circuit 20 as shown in "b" of FIG. 14, and is one-sided envelope waveform by the low-pass filter 21 ("c" in FIG. 14)
The envelope waveform is subjected to two differentiating processes using a first-order differentiating circuit 22 and a second-order differentiating circuit 23, and its zero cross point (envelope inflection point of the envelope waveform) is extracted by a Tg comparator 24. And the group arrival time tg. However, in order to extract only the zero cross point, which is the inflection point of the envelope waveform, the window signal for tg detection (see “d” in FIG. 14) by the Tg detection gate circuit 26 based on the magnitude of the first differential waveform. Dashed line) and set Tg with multiplier 28
The first rising of the product of the output of the comparator 24 and the output is defined as tg (“h” in FIG. 14). Similarly, for the tp signal, a window signal for tp detection (broken line in FIG. 14E) is set by the Tp gate circuit 27 based on the tg detection signal, and the multiplier 29 outputs the Tg comparator 25 output to the sensor 10 output signal. With the first fall
tp. Next, a method of calculating the distance γ between the input position of the pen 9 and the sensors 10a to 10c in the case of using the above detection method will be described. If the group velocities vg are equal, it can be obtained by γ = (vp * tp) + (offset value). When the group velocity is (vg ≠ vp), the phase in the group shifts with the distance γ. Therefore, tp is 15th
Step-like objects as shown in the figure are observed, and conversion work is required. This step-like shape indicates the movement of the phase detection point, and the connection between the stages is parallel movement by the signal period T. On the other hand, when the group velocity is equal to (vg = vp) and a constant phase detection point can always be observed, such a step cannot be performed, and a tp like a straight line a is obtained. Therefore, tp obtained in a stepwise manner in the case of (vg ≠ vp) is
What is necessary is just to convert to the original straight line. That is, tpa ≒ (vg / vp) tg−tof where (tof: offset value), but since tg has large fluctuation, tpi = nT ≒ (vg / vp) tg−tof−tp where (n: integer value Tpa = tp + T * Int [｛(vg / vp) tg-tof-tp｝ / T + 0.
5] is obtained as tpa, and γ is eventually calculated from the following equation. γ = vp * tpa-rof where (rof: offset value)

[Problems to be solved by the invention]

However, the pen 9 and the sensor 10
In order to accurately calculate the distance between a and 10c, it is assumed that an accurate calculation constant is used. However, due to the nature of the method of utilizing the propagation of ultrasonic waves, it is inevitable that such calculation constants will fluctuate slightly due to various physical constants and adjustment conditions of each product. However, it is difficult to mass-produce products with stable coordinate accuracy. For this reason, it is necessary to detect the difference between the detection points of tp in order to define the correct calculation constants for the sampling data (tg, tp), respectively. When sampling data from the entire effective area, for example, As shown in Fig. 16, it was necessary to input a large amount of continuous data.

[Problems to be solved by the invention]

The present invention has been made in order to solve the above-described problems, and has various physical constants of individual products, coordinate input with accurate coordinate accuracy regardless of subtle variations due to adjustment conditions and the like. It is intended to provide a device. The present invention has the following configurations to solve the above-mentioned problems. That is, in a coordinate input device that emits ultrasonic waves having different group velocities and phase velocities by using a coordinate pointing tool and measures the propagation time differences that reach at least three sensors installed at various locations of the propagation body to recognize the indicated coordinates. The coordinate input effective area is divided into a predetermined number, a continuous curve having an appropriate length is designated and input in the divided area, and an average value of an integer error distribution between the designated input position and the detection data of the input position is calculated. A first distance measuring means for obtaining a distance between the coordinate pointing device and the sensor based on the obtained integer conversion error value;
The input coordinates are calculated and output based on the difference between the distances measured by each sensor of the first distance measuring means. Further, the coordinate input effective area is divided into a predetermined number, a plurality of points in the divided area are designated and input, and an error between an integer value of one point in the designated input and an integer value of another point is obtained,
A second distance measuring means for calculating a distance between the coordinate pointing device and the sensor based on the obtained integer error value;
The input coordinates are calculated and output based on the distance measured by each sensor of the distance measuring means. Further, the first distance measuring means or the second distance measuring means calculates f in the integer value calculation expression based on the obtained integer error value.
Two constants of (vg / vp) and f, and, if necessary, the offset value are re-determined by the least squares method, and the input coordinates are calculated by using the constants based on the difference between the measured distances of the sensors.

[Action]

With the above configuration, the group arrival time tg and the phase arrival time tp are sampled at arbitrary plural points in the coordinate input effective area, and based on this, an optimum constant for each product is estimated, and each constant is estimated using the constant. The input coordinates are calculated based on the difference between the distances measured by the sensors.

【Example】

Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram of a coordinate input device according to one embodiment of the present invention. In FIG. 1, the same components as those in FIG. In the figure, reference numeral 1 denotes a pen driving circuit which drives the pen 9 and outputs the driving timing of the pen 9 to the distance measuring circuit 3;
The inflection point detection circuit 3 for detecting an inflection point from the detection signals from 10a to 10c detects the time difference between the inflection point detection timing in the inflection point detection circuit and the pen 9 driving timing from the pen driving circuit 1. A distance measuring circuit for measuring the distance between the input position of the pen 9 and each of the sensors 10a to 10c under control according to a control flow chart described later, and 4 is an input position of the pen 9 based on a distance measured by the distance measuring circuit 3. This is an input coordinate position output circuit that calculates and outputs coordinates. In the present embodiment, they are arranged on the four sides of the propagation body 11, respectively. The coordinate input device of this embodiment also has the appearance shown in FIG. 9 described above, and the inflection point detection circuit 2 has the configuration shown in FIG. The pen drive circuit 1 causes the pen 9 to vibrate, and the pen 9 generates an ultrasonic signal having a different propagation speed between a longitudinal wave and a transverse wave. Also in this case, the distance measurement circuit 3 performs the above-described operation in the same manner as in the related art: n = Int [{(vg / vp) tg−tp−tof} /T+0.5 (A)
It is necessary to perform an integer calculation called an expression. If the correct n value is not calculated here, the pen 9 and the sensor
As the distance γ between 10a and 10d, λ · Δn (= vp · T · Δn) where Δn: an error Δγ of n value error occurs, which has a significant effect on the coordinate accuracy. Therefore, the calculation constant of the n-value calculation is (vg / vp), tof, T
Must be set sufficiently reasonable. However, these constants fluctuate slightly due to the physical constants of individual products, the degree of adjustment, and the like, and are not constant. In the coordinate input device, a case where the pen 9 scans a straight line passing through the sensor 10d of the propagation body 11 as shown in FIG. 2 will be described as an example. In this case, if the correct (vg / vp), tof, T is used (when set correctly), the pen 9 and the sensors 10a to 10d are used.
The relationship between the distance between them and the integer error [(correct n value)-{(vg / vp) tg-tp-tof} / T)] is as shown in FIG. 3 (a). On the other hand, when (vg / vp) is set to be larger than an accurate value, the relationship between the distance between the pen 9 and each of the sensors 10a to 10d and the integer error is as shown in FIG. 3 (b). Similarly, for tof, T, FIGS. 3 (c) and 3 (d)
The distribution is as shown in FIG. If the distance between the pen 9 and each of the sensors 10a to 10d and the integer error are as shown in FIGS. 3 (b) to 3 (d), the detection operation is unstable, and When the pen 9 is tilted too much, when the pen pressure is unstable, when the propagator 11 is strongly pressed down by hand, the margin for other disturbance factors becomes small, and this is a threshold value for n-value error occurrence. A case exceeding ± 0.5 may also occur. Therefore, (vg / vp), tof, T is optimized as much as possible,
The state must be as shown in FIG. 3 (a). For this reason, in the distance measuring circuit 3 of the present embodiment, (tg, tp) is sampled in advance at arbitrary plural points in the effective area. Then, (vg / vp), tof, and T are optimized based on the plurality of sampled values, and the state shown in FIG. 3A is maintained. Based on the above equation (A), n = ｛(vg / vp) tg−tp−to｝ / T = [｛(vg / vp) / T｝ tg] − (1 / T) tp− (tof Considering the formula (/ T), if there are three data points of (tg, tp, n), three constants of (vg / vp) / T, 1 / T, (tof) / T can be determined. it can. However, as described above, tg fluctuates more than tp, so (tg, tp, n) is distributed over the entire effective area, a large number of points are taken in, and statistical processing is performed to obtain (vg / vp) / It is necessary to optimize T, 1 / T, (tof) / T. Although (tg, tp) at each designated point can be easily taken in, the corresponding n value is completely unknown.
In the present embodiment, for each (tg, tp) input for constant optimization, control is performed such that a correct n value is estimated and an optimum calculation constant is set based on it. Hereinafter, the optimum calculation constant setting process in the distance measurement circuit 3 will be described with reference to the flowchart of FIG. First, in step S1, in order to obtain a proper constant over the entire effective area, each area (for example, nine areas 1 to 9 in FIG. 5) is shown in FIG. A line segment having an appropriate length is scanned (tracing), and (tg, tp) in nine areas is sampled. Sampling point n for effective area of propagator 11
Is set as n _1St (i) (when 20 points per area are captured, (i = 1 to 180). For example, the data (tg, tp) obtained by the sensor 10d by this sanampling is An example arranged in the sampling order is shown in the lower part (b) of Fig. 6. For each area, assuming that the first point shown in Fig. 5 is n = 0, the detection point of tp is determined during scanning. In the case of a change, the constant initial value by error distribution calculation is calculated as +1 or -1 for (| tp _k -tp _k-1 | ≧ T). The initial state of the n-value is shown in the upper part (a) of Fig. 6. Therefore, the initial state of each n-value is determined by the area set in the following steps S2 and S3. Calculate the error average n _AVG (p) of (P = 1 to 9 when there are 9 areas): (vg / vp) / T, 1 Using / T, (tof) / T, Is calculated, the result shown in the lower part of FIG. 7 is obtained. The stairs of the error distribution indicate the difference between the set n value of each area and the true n value, but the average value of the error in each area is shifted by a certain numerical value between the areas. Therefore, based on the area which is the step S4, the n value of the other area is corrected by the value obtained by converting the difference between the average values into an integer. For example, the _{correction is made} as n (i) = n _1sT (i) + n _AVG (i). With reference to area 1, area 2 corrects all n values to +15, area 2 to +23,... Area 9 to +23. Then, the process proceeds to step S5 to check whether the error distribution range is within 1 or less. If the error distribution range becomes 1 or more here, it is considered that the optimum constant is abnormally shifted from the initially determined value.
To perform an additional error process. On the other hand, if the error distribution range falls within the level 1 or less, it means that the set n values have been corrected so as to be shifted by the same value in the processing of step S4, and the processing of step S10 and lower than step S5. Proceed to. An example in which the error distribution range after the correction of the n value falls within the level 1 or lower is shown in the upper part of FIG. In step S10, it was determined by the following procedure (tg, tp, n)
The optimal (vg / vp) T, 1 / T, (tof) / T for expressing the data group by the formula (A) is determined by statistical processing such as the least square method. Then, in the next step S11, the error distribution range is checked again, and it is checked whether or not the error distribution range is within a certain range, for example, whether or not the level is 2 or less. Here, for example, if the level is equal to or higher than 2, the process proceeds to the error processing in step S6. On the other hand, if the error distribution range is within level 2 or the like, the process proceeds from step S11 to step S12. Among the determined constants, top / T is the constant part. Here, the integer part (tof / T) _int is shifted from the true value by a certain value, but the decimal part (top / T) _dec indicates the true value. Therefore, n is calculated by the following equation, and (tof /
T) _int is included in γof in the γ calculation formula. By the way, tp has a sufficiently small fluctuation compared to tg and good linearity (the distance between the pen 9 and the sensors 10a to 10d). From, vp, γof can be easily obtained. This is easily obtained by γ = vp * tp−γof formula (B). Regarding the constants optimized by the above processing, it is desirable to output the result to the outside by some method using the input coordinate position output circuit 4 or the like thereafter. By performing such control, calibration can be performed based on this constant, such as by changing the setting of the deep switch. or,
A backup RAM may be provided and update recording may be performed here. As described above, according to the present embodiment, since the error distribution is averaged for each area and the n value is corrected, even if the error distribution range is large in each area to some extent, the average value is generally adjusted between the areas. Since the difference has a value close to the numerical value, a remarkable operation and effect that corrects the n value can be obtained. In this way, it is only necessary to continuously take in an appropriate amount from each of the dispersed areas (the amount that has a statistical averaging effect), and the calibration time can be shortened and simplified. In this embodiment, since the sampling data is greatly reduced as described above, it is easy to cause a microcomputer in the apparatus to execute a series of calibration processes. However,
It is of course possible to output (tg, tp) data to an external computer and to leave the subsequent calculation processing to the external computer. After the constant n is obtained as described above, the sensor, for example, 4 based on the input position of the pen 9 based on the constant n
The distance to the sensors 10a to 10d can be determined accurately. Then, the designated coordinate position with the pen 9 can be uniquely and accurately obtained from the distance between each of the sensors 10a to 10d and the input position of the pen 9. The calculation control of this position is known, and therefore, detailed description is omitted. Then, the input coordinates detected by the input coordinate position output circuit 4 are output to the connection processing device. As described above, according to this embodiment, each (t) sampled for calibration is calculated using a preset calculation constant.
g, tp), the correct constant n value is estimated, and then, based on the (tg, tp, n) data group created by the constant n, a more optimal calculation constant is calculated by a method such as the least square method. Determined by
Can be calibrated. For this reason, an accurate input coordinate well can be specified with a very small sampling amount.

[Other embodiments]

In the above-described first embodiment, a method of averaging the error distribution of each area and detecting the difference is used to detect the difference between the integer values between the areas. This allows
Even if the error distribution range in each area is large to some extent, n
There is a feature that the value can be estimated correctly. However, the present invention is not limited to the above example, and the erroneous coordinate distribution range in each area is sufficiently small.
(Difference between all errors) is (difference between correct n values) and ± 0.
If it is within the range of the difference of 5, n
The value can be estimated almost correctly. That is, this is the case where the following equation is satisfied. Int {Ern (i) -Ern (j) +0.5} = N (i) -N (j) i, j: All data combinations In this case, for (vg / vp) / T and 1 / T, it means that the difference between the actual value and the initially set value is within a predetermined range. That is, it is effective when a large variation in the constant is particularly limited to (tof) / T. Thus, the present invention in the case where the erroneous coordinate distribution range in each area is sufficiently small and (the difference between all errors) is within the range of (the difference between correct n values) and ± 0.5. Another embodiment according to the present invention will be described below with reference to the flowchart of FIG. The other configuration, for example, the hardware configuration is the same as that of the above-described first embodiment, and the description is omitted. First, in step S21, the propagation body 11 is divided into nine areas shown in FIG. 5, and in each area, the input of the sampling data n with the pen 9 is performed, for example, the sampling point n with respect to the effective area of the propagation body 11 is set to n _1St Set as shown in (i). (If 20 points are captured in one area, (i
= 1 to 180). At this time, in this embodiment, the n value is not determined at all by detecting the movement of the tp detection point as in the first embodiment. Therefore, it is not necessary to input continuous data even in each area, and data may be arbitrarily sampled whether continuous or discontinuous in each area. Next, the n value is set in step S22. First, for example, “1” is set for the first data. Then, for the second and subsequent data, the n value is set according to the following equation. In the following step S23, the above control is set (tg, tp,
For the data group of n), constant optimization is performed by a known least squares method. Then, using the constant optimized in step S24, the error distribution is checked according to the following equation. If the error distribution is not within the predetermined range, the process proceeds to step S26 as an abnormality, and error processing is performed. On the other hand, if the error distribution is within the predetermined range, it is judged as pass,
The configuration of constants is performed in the same manner as in step S12. After that, the input coordinates may be calculated and output in the same manner as in the first embodiment. As described above, according to the present embodiment, even if the propagation characteristics of the propagation body 11 and the pen 9 vary, the pen 9 and the propagation body
It is possible to independently and easily re-determine and calibrate the optimal calculation constant as a coordinate input device composed of 11 combinations. In addition, the data sampling amount at that time is greatly reduced as compared with the conventional case. This also greatly improves mass productivity. By performing control as described above, the constant n can be reliably obtained with simple control. As described above, according to each of the above-described embodiments, a correct constant n value is estimated for each (tg, tp) sampled for calibration using a preset calculation constant, and then this constant is Based on the (tg, tp, n) data group created by n, a more optimal calculation constant can be determined by a process such as the least squares method and calibrated. Furthermore, even if the propagation characteristics of the propagation body 11 and the pen 9 vary, if the dispersion is within a predetermined range, the optimal calculation as a coordinate input device composed of a combination of the pen and the propagation body is performed. The constants can be independently and easily re-determined and calibrated, and the amount of data sampling at that time is greatly reduced as compared with the conventional case. This also greatly improves mass productivity.

【The invention's effect】

As described above, according to the present invention, it is possible to suppress the influence of the variation in the propagation characteristics of the propagation body and the pen with a small sampling amount, and to independently and easily redetermine the optimal calculation constant as the coordinate input device. And can be calibrated.

[Brief description of the drawings]

FIG. 1 is a block diagram of a coordinate input device according to one embodiment of the present invention, FIG. 2 is a diagram showing an example of sampling data input in this embodiment, FIG. 3 is a diagram showing an example of an integer error distribution, FIG. FIG. 5 is a flowchart showing the constant calibration mode processing of this embodiment. FIG. 5 is a diagram showing each sampling area in this embodiment. FIG. 6 shows an example of inflection point detection in each sampling area shown in FIG. FIG. 7 is a diagram showing an integer error distribution after error correction of the sensor detection waveform in each sampling area shown in FIG. 5, and FIG. 8 is a diagram showing a constant calibration mode process of another embodiment according to the present invention. FIG. 9 is a schematic external view of a coordinate input device, FIG. 10 is a diagram showing a pen driving waveform, FIG. 11 is a diagram showing a detection output waveform from a sensor, and FIG. 12 is a detection output waveform from a sensor. Peak position detector Block diagram, FIG. 13 is a block diagram showing another configuration of the peak position detecting section of the detection output waveform from the sensor, FIG. 14 is an operation timing chart of each part in FIG. 13, FIG. 15 between the pen and the sensor. FIG. 16 is a diagram showing an example of inputting conventional sampling data. In the figure, 1... Pen drive circuit, 2... Inflection point detection circuit, 3
... Distance measuring circuit, 3a ... Constant setting section, 4 ... Input coordinate position output circuit, 9 ... Pen, 10 ... Sensor, 11 ... Propagator, 12 ... Vibration absorber, 13 ... First-stage amplifier , 14 ... Full-wave rectifier circuit, 15,21 ... Low-pass filter, 16 ... Differential circuit, 1
7 Comparator, 19 Preamplifier, 20 Absolute value circuit, 22 Primary differential circuit, 23 Secondary differential circuit, 24
... Tg comparator, 25... Tg comparator, 26... Tg gate circuit, 27.

Claims (5)

(57) [Claims] 1. Coordinates which emit ultrasonic waves having different group velocities and phase velocities by means of a coordinate pointing device and measure the propagation times at which they reach at least three sensors installed at various locations of the propagation body, thereby recognizing the indicated coordinates. In the input device, the coordinate input effective area is divided into a predetermined number, a continuous curve having an appropriate length is designated and input in the divided area, and an integral error distribution between the designated input position and the detection data of the input position is input. An average value is obtained, and first distance measuring means for obtaining a distance between the coordinate pointing device and the sensor based on the obtained integer error value is provided. The distance measured by each sensor of the first distance measuring means is provided. A coordinate input device, which calculates and outputs input coordinates according to (1). 2. The first distance measuring means re-determines three constants of f (vg / vp), f and an offset value in an integer value calculating formula by a least square method based on the obtained integer error value. And
2. The coordinate input device according to claim 1, wherein input coordinates based on a difference between measured distances of the sensors are calculated using the constant. 3. Coordinates which emit ultrasonic waves having different group velocities and phase velocities by means of a coordinate pointing device and measure the propagation times at which they respectively reach at least three sensors installed at various places of the propagation body to recognize the indicated coordinates. In the input device, the coordinate input effective area is divided into a predetermined number, a plurality of points in the divided area are designated and input, and an error between an integer value of one point in the designated input and an integer value of another point is calculated. And a second distance measuring means for calculating a distance between the coordinate pointing device and the sensor based on the obtained integer error value. The input coordinates are determined by a distance measured by each sensor of the second distance measuring means. A coordinate input device for calculating and outputting. 4. The second distance measuring means calculates two values of f (vg / vp) and f in an integer value calculation formula based on the obtained integer error value.
2. The coordinate input device according to claim 1, wherein the two constants are re-determined by the least squares method, and the input coordinates based on the difference between the measured distances of the sensors are calculated using the constants. 5. Coordinates which emit ultrasonic waves having different group velocities and phase velocities by means of a coordinate pointing device, and measure the propagation times to reach at least three sensors installed at various places of the propagation body to recognize the indicated coordinates. A coordinate input method in an input device, wherein a coordinate input effective area is divided into a predetermined number, a continuous curve having an appropriate length is designated and input in the divided area, and the designated input position and input position detection data are input. The average value of the integer error distribution is obtained, the distance between the coordinate pointing device and the sensor is further obtained based on the obtained integer error value, and the input coordinates are calculated and output based on the measured distance at each sensor. A coordinate input method characterized by being made possible.