CN1567165A - Coordinate predicting and estimating filtration method of touch control panel - Google Patents

Abstract

This invention discloses a touch control panel coordinate prediction and prediction filtering method which is to use a filter with prediction and prediction filtering in the touch control panel and to predict the X, Y coordinate and prediction filtering process. The method are the following: first to predict prior X,Y coordinates by a filter; to calculate and sample circle optimism prediction value by prediction technique; to predict the optimism X, Y coordinates by orthogonality principle between the measured X,Y coordinate and predicted X,Y coordinate; to get a more accurate coordinates and a better correlation between front and back coordinate position.

Description

Touch control panel coordinate is estimated and is estimated filtering method

Technical field

The present invention relates to a kind of touch control panel coordinate estimates and estimates filtering method, especially refer to a kind ofly estimate and estimate wave filter through one, X, Y coordinate signal are estimated and estimate Filtering Processing, made resulting X, the Y coordinate is more accurate and front and back coordinate position correlativity better method.

Background technology

Contact panel (Touch Panel) is a kind of hommization input media.Because input media in the past, for example keyboard, mouse, trace ball etc. all can't satisfy the user and import demand easily, therefore input is convenient to combine multiple input media to industry in order to ask, to improve input efficiency and convenience, but along with design of electronic products all with light, thin, short, little, function is intrepid is developing direction, input medias such as conventional keyboard can't satisfy this demand, also do not have the space on the product and hold multiple input media.And contact panel also has the operating function of hommizations such as keyboard, mouse function and handwriting input simultaneously except meeting option designing requirement at many levels, makes contact panel become the optimal selection of man-machine interface.

Contact panel mainly is made of the face glass that transparent conducting glass and transparent conductive film constitute, and it is by the control circuit board control of being made up of the integrated circuit control chip, shows that with touch control manner required image is on screen.And contact panel can be divided into resistance-type, condenser type, electro-induction formula etc. according to the induction mode difference, is wherein the most extensively used with resistance-type again.And the contact panel of this resistance-type can be divided into four-wire type and five-line etc. according to the wiring mode difference, its know-why be with transparent conducting glass and transparent conductive film respectively according to X, Y-axis wiring, and between the electro-conductive glass of the conductive film on upper strata and lower floor, be provided with and strut space invocation point partition.When finger, pointer or other medium are exerted pressure to the conductive film on upper strata, the levels electrode is communicated and produce potential difference (PD), promptly form a switch, push and promptly produce the ON/OFF effect, so that a signal to be provided, this signal is passed to microprocessor controller through winding displacement again and is handled, and the coordinate position at the further calculating place of exerting pressure also is shown on the screen.

Yet, no matter the induction mode of this contact panel is resistance-type or condenser type or electro-induction formula, its purpose is all X, the Y coordinate on the detecting contact panel, and the X of its gained, Y coordinate figure all have noise, and makes X, the Y coordinate and actual X, Y coordinate generation error of detecting.

So, how to improve traditional contact panel because of noise produces the disappearance of error, be emphasis of the present invention.

Summary of the invention

In view of the foregoing, the purpose of this invention is to provide a kind of touch control panel coordinate estimates and estimates filtering method.

For achieving the above object, the present invention is by the following technical solutions: a kind of touch control panel coordinate is estimated and is estimated filtering method, it is to utilize one to have and estimate and estimate filter filtering in a contact panel, to X, the Y coordinate signal is estimated and is estimated Filtering Processing, comprising the following step: with this wave filter earlier to previous X, the Y coordinate is estimated, calculate best discreet value of a sample period with pre-estimating technology, again by present measured X, Y coordinate and the preceding X that estimates, the Y coordinate is estimated the X that the best with the quadrature principle, the Y coordinate makes the estimation X that obtains, the Y coordinate is more accurate and better with front and back coordinate position correlativity.

Described wave filter is a Kalman filter.

Description of drawings

Fig. 1 is an action flow chart of the present invention

Fig. 2 for the present invention through the resulting X of Kalman filter, the Y coordinate is more accurate and with the better synoptic diagram of front and back coordinate position correlativity

Embodiment

See also Figure 1 and Figure 2, the present invention estimates for a kind of touch control panel coordinate and estimates filtering method.In the present embodiment, this contact panel is a resistance-type five line contact panels 1, and this contact panel 1 more includes an analog-digital converter 2, a microprocessor 3 and has the Kalman filter 4 (Kalman Filter) of estimating and estimating filtering.When this contact panel 1 of contact, this contact point is in the X of this contact panel 1, the Y coordinate respectively spreads out of an analog voltage signal, this signal converts 12 digital signal to through analog-digital converter 2, import microprocessor 3 again into and be construed to X on the contact panel 1, the Y coordinate position, the X of gained, the Y coordinate position, again by Kalman filter 4, earlier to previous X, the Y coordinate is estimated, calculate best discreet value of a sample period with pre-estimating technology, again by the present X that measures, Y coordinate and the preceding X that estimates, the Y coordinate is estimated the X that the best with quadrature principle (Orthogonal Principle), the Y coordinate is for the estimation X that obtains, the Y coordinate is more accurate and better with front and back coordinate position correlativity.

Its expression formula is as follows:

The signal of supposing X, Y coordinate gained is Xdata, Ydata, and its linear equation is

Z (k)=θ m (k)+vm (k), order $\mathrm{\θm}\left(k\right)=\left[\begin{array}{c}\mathrm{Xdata}\left(k\right)\\ \mathrm{Ydata}\left(k\right)\end{array}\right]$

Xdata (k), Ydata (k) that Z (k) measured for present stage, Vm (k) is tool mean value ε _m, and from variance δ _mAdditive white Gaussian.

(1), pre-estimation formula

$\widehat{\mathrm{\θ}}m\left(k\right|k-1)=A^{*}\widehat{\mathrm{\θ}}m(k-1|k-1)$

P(k|k-1)＝A ^TP(k-1|k-1)A+ωm(k-1)

In the formula $A=\left[\begin{array}{cc}1& 0\\ 0& 1\end{array}\right],$ Suppose $P\left(0\right|0)=\left[\begin{array}{cc}1& 0\\ 0& 1\end{array}\right],$ $\mathrm{\ωm}\left(0\right)=\left[\begin{array}{cc}\mathrm{qm}& 0\\ 0& \mathrm{qm}\end{array}\right],$ P (k) is for estimating the factor,

Wm (k) is for estimating the noise variance number

(2) estimation calculating formula:

K(k)＝P(k|k-1)C ^T[CP(k|k-1)C ^T+vm(k)] ^-1

$\widehat{\mathrm{\θ}}m\left(k\right|k)=A\widehat{\mathrm{\θ}}\left(k\right|k-1)+K\left(k\right)[Z\left(k\right)-C\widehat{\mathrm{\θ}}m\left(k\right|k-1)]$

P(k|k)＝[1-K(k)C]P(k|k-1)

In the formula $\mathrm{vm}\left(k\right)=\left[\begin{array}{cc}\mathrm{\δm}& 0\\ 0& \mathrm{\δm}\end{array}\right],$ C＝1， $I=\left[\begin{array}{cc}1& 0\\ 0& 1\end{array}\right],$ K (k) is the estimation factor

Above calculating formula finish Z with Signal fused, estimate best X, Y coordinate figure $\hat{Z}\left(k\right)=\widehat{\mathrm{\θ}}m\left(k\right|k).$

When present embodiment is implemented in reality, as shown in Figure 2, utilized the Matlab software emulation, design standardized sin ripple on contact panel, X, Y coordinate x, the y equation of setting the sin ripple are as follows:

X is by 0 to 6.28, and each at interval 0.1; Totally 62 of X, Y coordinate

x＝0∶0.1∶2*3.14；

y＝sin(x)；

Parameter setting Pk (1)=1; Wm=0.1; Vm=0.1; The factor is estimated in the Pk representative; Wm represents estimation error; Vm represents noise from variance

Suppose X, Y coordinate that contact panel is detected, change into digital signal, import microprocessor 3 again into and be interpreted as X on the contact panel 1, Y coordinate position Xdata, Ydata through analog-digital converter 2, had noise, vx is an X coordinate noise, and vy is the noise of Y coordinate, and expression is as follows:

Because of totally 62 on X, Y coordinate

Work as i=1: 1: 62

The noise of supposing the X coordinate is ± 0.2; The noise of Y coordinate is ± 0.2

vx(i)＝(-1)^i*rand(1)/5；

vy(i)＝(-1)^i*rand(1)/5；

Xdata(i)＝x(i)+vx(i)；

Ydata(i)＝y(i)+vy(i)；

Utilize again that Kalman filter 4 is made Xdata, Ydata estimates, at first carry out pre-estimating technology and calculate best discreet value Xt of a sample period, Yt, then reading in new measured value is Xtt, Ytt with quadrature principle renewal Xdata, Ydata estimated value, with SNR_m is expression Xdata, the margin of error of Ydata and x, y, SNR_p is expression Xtt, the margin of error of Ytt and x, y, and program is as follows: (wherein Kk is the estimation factor)

The initial data of X equals survey data

Xtt(1)＝Xdata(1)；

The initial data of Y equals survey data

Ytt(1)＝Ydata(1)；

If survey data and real data initial error value are 0

SNR_m＝0；

If estimation data and real data initial error value are 0

SNR_p＝0；

Begin to estimate by second document

Work as i=2: 1: 62

Xt is estimated by Xdata

Xt(i)＝Xdata(i-1)；

Yt is estimated by Ydata

Yt(i)＝Ydata(i-1)；

Estimate the factor and estimate the factor with wm calculating and get by last pen

Pk(i)＝Pk(i-1)+wm；

Estimation factor calculating formula

Kk(i)＝Pk(i)/(Pk(i)+vm)；

X coordinate estimation factor calculating formula

Xtt(i)＝Xt(i)+Kk(i)*(Xdata(i)-Xt(i))；

Y coordinate estimation factor calculating formula

Ytt(i)＝Yt(i)+Kk(i)*(Ydata(i)-Yt(i))；

Upgrade Pk (i) value

Pk(i)＝(1-Kk(i))*Pk(i)；

Calculate the error amount of metrology data and real data

SNR1＝(Xdata(i)-x(i))^2+(Ydata(i)-y(i))^2；

Calculate the error amount of estimation data and real data

SNR2＝(Xtt(i)-x(i))^2+(Ytt(i)-y(i))^2；

All measurement error value

SNR_m＝SNR_m+SNR1；

All estimation error values

SNR_p＝SNR_p+SNR2；

Simulation result as shown in Figure 2, Xtt, Ytt be data after Kalman filter 4 estimation near the sin ripple, and SNR_p=0.3645＜SNR_m=1.5687, expression Xtt, Ytt have the less margin of error.

In sum, the present invention has the Kalman filter of estimating and estimating filtering with one, and X, Y coordinate signal are estimated and estimated Filtering Processing, make X, the Y coordinate is more accurate and better with front and back coordinate position correlativity.

Claims (2)

1. a touch control panel coordinate is estimated and is estimated filtering method, and it is to utilize one to have and estimate and estimate filter filtering in a contact panel, and X, Y coordinate signal are estimated and estimated Filtering Processing, comprising the following step:

Earlier previous X, Y coordinate are estimated with this wave filter, calculate best discreet value of a sample period with pre-estimating technology, estimate X, the Y coordinate that the best by now measured X, Y coordinate and the preceding X that estimates, Y coordinate with the quadrature principle again, make the estimation X, the Y coordinate that obtain more accurate and better with front and back coordinate position correlativity.

2. estimate and estimate filtering method according to claims 1 described touch control panel coordinate, it is characterized in that: described wave filter is a Kalman filter.

Images