CN101807110A - Pupil positioning method and system - Google Patents

Abstract

A pupil location method and system, the method comprising the steps of: capturing an eye image of a user, and positioning a pupil center and a light reflecting point by using the eye image; defining a distance value by the distance from the pupil center to the light reflecting point, and defining an angle value by the pupil center position and a reference line matched with a vector of the light reflecting point relative to the pupil center; and converting the position corresponding to the screen when any point of the screen is watched by utilizing an angle-distance distribution graph obtained by watching the distance values and the angle values of different positions. The invention does not need to wear additional equipment on the human body and does not need complex calculation, and can more accurately and effectively control the positioning function of the computer pointer.

Description

Pupil positioning method and system

technical field

The invention relates to a pupil positioning method and system, in particular to a pupil positioning method and system which can control a computer with eyes.

Background technique

At present, eye tracking technology can be divided into contact type and non-contact type according to whether it is in contact with human eyes. Contact eye tracking technology can be divided into search coil method and electrooculogram method. Non-contact eye tracking technology mainly Based on vision recognition, it can be divided into head-mount or free-head.

In terms of contact eye tracking technology, the search coil method allows the user to wear a soft lens with an induction coil. When the user turns the eyeball to drive the lens, the induction coil will generate an induced electromotive force due to the change of the magnetic flux. The magnitude of the electromotive force represents the angle of eyeball deflection, but the disadvantage of this method is that it is easily affected by the user's eyeball conditions, such as eye secretions, etc., and the soft lens has a double-layer structure, which will affect the user's vision; as for The electrooculogram (EOG) method is to attach a large number of electrodes around the eyes, and use the electrodes to detect the voltage difference generated by the eyeball rotation to judge its angle of up, down, left, and right. The disadvantage is that the skin of the face where the electrodes are attached Resistors tend to get unstable electrical signals due to horny secretion, and can only record the giant turning of the eyeball, but cannot record smaller angle changes.

In terms of head-mounted eye tracking technology, the user must wear a pair of glasses with a small camera. Since the relative distance between the eyes and the camera is fixed, it will not be caused by face deviation or changes in the relative distance of the eyes. The judgment is inaccurate, so when the user uses the glasses, the glasses must be fixed on the head so as to fix the relative position between the small camera and the eyes, which is not only inconvenient but also uncomfortable for the user.

In terms of head-mounted eye-tracking technology, there are eye trackers with screens and dual CCD cameras in foreign countries. The more famous ones in Taiwan are related researches by Lin Chensheng and others using screens and single CCD cameras. However, the aforementioned head-mounted eye-tracking technology uses relatively complex calculations. In addition, although the eye tracker with dual CCD cameras can accurately locate the target, it is very expensive and requires two CCD cameras.

The lack of eye control technology in the past is that no matter when it is implemented in contact or non-contact, precise positioning is required, and precise positioning requires expensive hardware and software equipment. As a result, eye control technology cannot be popularized and the general public can also be used.

Contents of the invention

In order to overcome the problem that the existing eye control technology needs expensive equipment to have high precision, the present invention provides a pupil positioning method that can accurately locate the pupil only by program judgment.

Therefore, the technical solution adopted by the pupil positioning method of the present invention to solve its technical problems is to use a pupil positioning method that can be realized by program software. The method is to capture the user's eye image when a user is watching a screen , and use the eye image to locate a pupil center and a reflective point; a distance value is defined by the distance from the pupil center to the reflective point, and a reference line is defined by the pupil center position and a vector of the reflective point relative to the pupil center is defined Obtain an angle value; use the angle-distance distribution map obtained by viewing the distance value and angle value of different positions, and obtain a coordinate corresponding to the angle-distance distribution map by viewing the distance value and angle value of any position according to a predetermined method point, so as to convert the position corresponding to the screen when watching any point on the screen, and finally locate the position where the user watches the screen. The user only needs to install the computer program software of the method of the present invention for discrimination, and the purpose of high accuracy can be achieved.

In order to overcome the problem that the existing eye control technology requires expensive equipment, the present invention provides a pupil positioning system that can accurately locate only simple equipment.

Therefore, the technical solution adopted by the pupil positioning system of the present invention to solve its technical problems is to include a light source that emits a beam of light to irradiate a user's eye, a camera that captures the user's eye image, and a computing module; the computing module has A feature positioning unit, a distance-angle processing unit and a coordinate conversion unit.

The feature positioning unit obtains the eye image and locates a pupil center and a reflection point; the distance-angle processing unit defines a distance value based on the distance from the pupil center to the reflection point, and defines a reference line with the pupil center position to match the reflection point A vector at the center of the pupil defines an angle value; the coordinate conversion unit uses an angle-distance distribution map obtained by viewing the distance value and angle value of different positions, and then uses the distance value and angle value of any position according to a predetermined method A coordinate point is correspondingly obtained in the angle-distance distribution map, and the position corresponding to the screen when viewing any point on the screen is converted.

In this way, the user only needs to have the above-mentioned light source, camera and computing module, and does not need other expensive equipment, so the purpose of reducing the hardware cost can be achieved.

The overall beneficial effect of the present invention is that the cost can be reduced without wearing additional equipment on the human body, and the angle-distance distribution map obtained by viewing different positions is used to convert the position corresponding to the screen when viewing any point on the screen, so there is no need for complicated calculations. It can control the positioning function more accurately and effectively.

Description of drawings

Fig. 1 is a schematic diagram illustrating a preferred embodiment of the pupil positioning system of the present invention;

Fig. 2 is a schematic diagram illustrating a distance value and an angle value defined by the center of the pupil and the reflective point;

Fig. 3 is a flowchart illustrating a preferred embodiment of the pupil location method of the present invention;

Fig. 4 is a flow chart illustrating the flow process of the training stage in the embodiment of the grouping correspondence method used by the pupil positioning system of the present invention;

Fig. 5 is a schematic diagram illustrating that the screen is divided into multiple groups of regions, but only one group of regions is displayed at a time;

Fig. 6 is a schematic diagram illustrating the positioning method of the pupil center and the reflective point;

Fig. 7 is a schematic diagram illustrating the distance-angle difference between the distance value and the angle value and each group area of the screen;

Fig. 8 is a schematic diagram illustrating the distance-angle distribution map of areas "1" to "16", and a coordinate point can be obtained correspondingly in the angle-distance distribution map with the distance value Li and the angle value θi;

Fig. 9 is a flowchart illustrating the flow process of the application stage in the embodiment of the grouping correspondence method used by the pupil positioning system of the present invention;

Fig. 10 is a schematic diagram illustrating that the distance-angle point set corresponding to each area will obtain an irregular fan-shaped distribution;

Fig. 11 is a schematic diagram illustrating that a sector drawn with a small number of distance-angle points can be converted into a relatively regular rectangle by using a twisting operation, and then the position of the rectangle is corresponding to the coordinates of the screen;

Fig. 12 is a flowchart illustrating the flow process of the application stage in the embodiment of the interpolation correspondence method used by the pupil positioning system of the present invention;

Figure 13 is a flowchart illustrating the flow of the application phase of an embodiment of the interpolation correspondence method used by the pupil location system of the present invention.

Detailed ways

Below in conjunction with accompanying drawing and embodiment the present invention is described in detail:

Before the present invention is described in detail, it should be noted that in the following description, similar components are denoted by the same numerals.

1. System architecture and method concepts

Referring to Fig. 1, the preferred embodiment of pupil location system 1 of the present invention comprises a light source 11, a video camera 12, an operation module 13 and a screen 14; Operation module 13 includes a training unit 130, a feature location unit 131, a distance -An angle processing unit 132 and a coordinate converting unit 133 .

In this preferred embodiment, the calculation module 13 is a computer program recorded on a storage medium, used to make the computer execute the pupil positioning method of the present invention, and the method is implemented according to the following steps:

1 to 3, the training unit 130 displays a target area on the screen 14 for the user to watch at a time, and then displays another target area for the user to watch after step 301 and step 302 are executed continuously until a predetermined number of targets are reached. The regions are all displayed; the feature positioning unit 131 obtains an eye image 401 of a user 4, and locates a pupil center 41 and a reflection point 42 (step 301), and the eye image 401 is taken by the camera 12, and the reflection point 42 It is obtained by using the light source 11 to emit a light beam to irradiate human eyes and reflect it.

Then, the distance-angle processing unit 132 defines a distance value L with the distance from the pupil center 41 to the reflective point 42, and defines a reference line (ie, the horizontal axis) with the position of the pupil center 41 to match the distance between the reflective point 42 and the pupil center 41. A vector 43 defines an angle value θ (step 302).

Then, the coordinate conversion unit 133 uses an angle-distance distribution map obtained by viewing the distance value L and the angle value θ of different positions (step 303). When the viewer watches any position of the screen 14, the distance value and the angle value of the position are correspondingly obtained in an angle-distance distribution diagram according to a predetermined method to obtain a coordinate point (step 304), so that when viewing any point of the screen 14, the Corresponds to the position of screen 14.

It should be noted that, when the present invention defines the angle value θ as a negative number, it is converted into an angle between 0 degrees and 180 degrees according to the following formula:

ifAngle<0then

Angle=180+Angle;

Two preferred embodiments of how to convert the distance value L and the angle value θ to the position of the pointer 101 corresponding to the screen 14 are introduced below. One of the predetermined methods is the grouping correspondence method, and the other is the interpolation correspondence method.

Two, the embodiment of group correspondence method

The pupil positioning method of the present invention can be divided into a training stage (as shown in Figure 4) and an application stage (as shown in Figure 9), and the flow process of each stage is described in detail as follows:

Referring to Figure 4, the process of the training phase includes the following steps:

1, the training unit 130 pre-divides the screen 14 into multiple groups of regions, and only one group of regions is displayed at a time, and each time a group of regions is displayed for the user to watch, the control feature positioning unit 131 performs positioning and controls distance-angle processing The unit 132 obtains the distance-angle of each group of regions; and if it is determined that all the regions have been displayed, obtains the distance-angle distribution map of each group of regions.

As shown in Figure 5, screen 14 is divided into 4x4=16 groups of regions, but only shows wherein one group of regions at every turn; An area I is displayed for the user to view (step 202).

Next, capture a user's eye image 401 (step 203), and use the eye image 401 to locate a pupil center 41 and a reflective point 42 (step 204); as shown in Figure 6, the location of the pupil center 41 The method is to calculate the position of the pupil center 41 after the eye image 401 is processed by dynamic binarization (Adaptive threshold) and Morphology (Morphology) closing (Close) operation.

Then, define a distance value L and an angle value θ (step 205) by using the center of the pupil and the reflection point; with the count value I+1 (step 206); then judge whether I=16 (step 207)? If not, continue to repeat steps 202 to 207 until it is judged that I=16, indicating that the distance-angle distribution diagrams of each group of areas have been obtained (step 208).

How to define the distance value L and the angle value θ can refer to the explanatory text of the above-mentioned FIG.

Referring to Fig. 7(a), the screen 14 first displays the "1" area, and now the distance value L1 and the angle value θ1 are obtained; referring to Fig. 7(b), the screen 14 displays the "2" area, and the distance value L2 and the angle value are obtained at this time Value θ2; Referring to Fig. 7 (c), screen 14 displays "3" area, and now distance value L3 and angle value θ3 are obtained; referring to Fig. 7 (d), screen 14 displays "4" area, and now distance value L4 is obtained and the angle value θ4; it is worth noting that since the eyesight moves horizontally, the distance values L1, L2, L3, and L4 will not differ too much, but the angle values θ1, θ2, θ3, and θ4 will have large changes.

But when the eye line of sight moves down to the next row of numbers "5", "6", "7" and "8" of "1", "2", "3" and "4", because the pupil moves down a little The distance, so that the distance value obtained when looking at "5", "6", "7" and "8" will become smaller, and the angle value will not have much difference from θ1, θ2, θ3 and θ4 respectively.

Referring to FIG. 8 , according to the aforementioned principle, and because the eyeball is a spherical surface rather than a plane, the distance-angle distribution diagrams such as areas "1" to "16" are obtained.

Referring to Fig. 9, and in conjunction with Fig. 6, the process of the group correspondence method in the actual application stage includes the following steps:

Capture a user's eye image 401 (step 211), and use the eye image 401 to locate a pupil center 41 and a reflective point 42 (step 212); use the pupil center and reflective point to define a distance value Li and An angle value θi (step 213); please refer to the description of FIG. 2 for the above steps, and will not repeat them here.

Referring to Fig. 8 again, a coordinate point 5 can be obtained correspondingly in the angle-distance distribution diagram with the distance value Li and the angle value θi, and the minimum objective function is used to find out the area closest to the coordinate point 5 (step 214), and thus it is completed The positioning function of the screen pointer (step 215).

The formula of the minimum objective function is as follows:

Min Obj.＝W1|Dist_t-Dis_i|+W2|Angle_t-Angle_i|

(i=1~16)

/Dist_i and Angle_i are the target distance value and angle value, Dis_t and Angle_t are the current distance value and angle value; W1 and W2 are distribution weights, this is to balance the contribution of the two/

$\mathrm{Dist}=0\sqrt{[{(x_1-x_2)}^2+{({\mathrm{the\; y}}_1-{\mathrm{the\; y}}_2)}^2]};$ /distance/ formula1

$\mathrm{angle}={\mathrm{the\; tan}}^{-1}\left[\frac{{\mathrm{the\; y}}_1-{\mathrm{the\; y}}_2}{x_1-x_2}\right];$ /angle/ formula 2

It can be seen from this preferred embodiment that this method uses the distance and angle of the pupil center and the reflective point obtained from multiple groups of areas to establish a corresponding relationship with the space of multiple groups of areas on the screen, and then obtain the positioning function effect ; However, the grouping correspondence method needs to look at more groups of areas, the better the positioning effect will be, so the interpolation correspondence method described below is used to improve this shortcoming.

Three, the embodiment of interpolation corresponding method

Referring to Figure 10, if the screen is divided into 8x8=64 groups of areas, the distance-angle point set corresponding to each area will get an irregular fan-shaped distribution; in this way, if you want to use the interpolation method to obtain the coordinates corresponding to the screen I encountered difficulties.

Referring to FIG. 11 , a sector 51 drawn with a small number of distance-angle points can be converted into a relatively regular rectangle 52 by using an Un-warping operation, and then the position of the rectangle 52 corresponds to the coordinates of the screen 14 , so that a fairly approximate and good corresponding effect can be obtained.

The pupil positioning method of the present embodiment is also divided into a training stage and an application stage, and the flow process of each stage is described in detail as follows:

Referring to Figure 12, the process of the training phase is similar to Figure 2, and also includes the following steps:

First, set the count value I=1 (step 601), then display the area I for the user to watch each time (step 602); then, capture a user's eye image 401 (step 603), and use the eye The first image 401 locates a pupil center 41 and a reflection point 42 (step 204); then, utilizes the pupil center and the reflection point to define a distance value L and an angle value θ (step 605); count value I+1( Step 606); judge again that I=16 (step 607)? If not, continue to repeat steps 02 to 07 until it is judged that I=16, indicating that the distance-angle distribution maps of each group of areas have been obtained (step 608 ).

The difference is that this method also performs a twisting operation on the aforementioned distance-angle distribution diagram to obtain a normalized distribution diagram (step 609 ), and corrects the normalized distribution diagram by using an affine transformation technique (step 610 ).

The normalized distance-angle distribution diagram obtained in step 609 has the following formula:

$\left\{\begin{array}{c}u=a_0+a_{1}x+a_2\mathrm{the\; y}+a_3{x}^2+a_4\mathrm{xy}+a_5{\mathrm{the\; y}}^2\\ v={\mathrm{<figure-callout\; id="0"\; label="b"\; filenames="H2009100067693E0000081.png"\; state="\{\{state\}\}">b</figure-callout>}}_0+b_{1}x+{\mathrm{<figure-callout\; id="2"\; label="b"\; filenames="H2009100067693E0000041.png,H2009100067693E0000061.png"\; state="\{\{state\}\}">b</figure-callout>}}_2\mathrm{the\; y}+b_3{x}^2+{\mathrm{<figure-callout\; id="4"\; label="b"\; filenames="H2009100067693E0000011.png,H2009100067693E0000041.png"\; state="\{\{state\}\}">b</figure-callout>}}_4\mathrm{xy}+{\mathrm{<figure-callout\; id="5"\; label="b"\; filenames="H2009100067693E0000061.png,H2009100067693E0000071.png"\; state="\{\{state\}\}">b</figure-callout>}}_5{\mathrm{the\; <figure-callout\; id="2"\; label="y"\; filenames="H2009100067693E0000041.png,H2009100067693E0000061.png"\; state="\{\{state\}\}">y</figure-callout>}}^2\end{array}\right.;$ Formula 3

$\left[\begin{array}{c}u\\ v\end{array}\right]=\left[\begin{array}{cccccc}{a}_0& a_1& a_2& a_3& a_4& a_5\\ b_0& b_1& b_2& b_3& b_4& {\mathrm{<figure-callout\; id="5"\; label="b"\; filenames="H2009100067693E0000061.png,H2009100067693E0000071.png"\; state="\{\{state\}\}">b</figure-callout>}}_5\end{array}\right]\left[\begin{array}{c}1\\ x\\ \mathrm{the\; y}\\ x^2\\ \mathrm{xy}\\ {\mathrm{the\; <figure-callout\; id="2"\; label="y"\; filenames="H2009100067693E0000041.png,H2009100067693E0000061.png"\; state="\{\{state\}\}">y</figure-callout>}}^2\end{array}\right]$ Formula 4

The correction operation is to convert the sampling point (X, Y)={(x ₁ , y ₁ ),..(x _n , y _n )} to the target point (X _T , Y _T )={ (x _T1 , y _T1 ),...(x _Tn , y _Yn )}, such as: input at least 6 sets of sample points (x ₁ , y ₁ ), (x ₂ , y ₂ ), ( x ₃ , y ₃ ), (x ₄ , y ₄ ), (x ₅ , y ₅ ) (x ₆ , y ₆ ) are turned into corresponding six groups of target points (u ₁ , v ₁ ), (u ₂ , v ₂ ), (u ₃ , v ₃ ), (u ₄ , v ₄ ), (u ₅ , v ₅ )(u ₆ , v ₆ ); where X, Y represent distance value Li, angle value θi, n= number of samples.

In order to obtain the optimal parameters of the second-order derivative function, the twisting operation in step 609 can obtain the optimal solution of 12 sets of coefficients of a ₀ ~ a ₅ , b ₀ ~ b ₅ through the operation of the inverse matrix; obtain a0 ~ After solving the 12 sets of coefficients of a5, b0~b5, the current unknown point (X, Y) is brought into the formula 4, and the value of the new coordinate point (u, v) can be obtained.

Step 610 is to use the Affine transform technology to perform moving calibration on the pupil position. The moving calibration is mainly to eliminate the influence of image scaling, image translation and image rotation on the coordinates, and then obtain a normalized coordinates.

$\left[\begin{array}{c}{x}^{\&prime;}\\ {\mathrm{the\; y}}^{\&prime;}\end{array}\right]=\left[\begin{array}{cc}a& b\\ c& d\end{array}\right]\left[\begin{array}{c}x\\ \mathrm{the\; y}\end{array}\right]+\left[\begin{array}{c}e\\ f\end{array}\right];$ Formula 5

where (x', y') are the new coordinates and a, b, c, d, e, f are the coefficients of the affine transformation. The calculation method of affine transformation is to input three pairs of coordinate points of any three corners of the screen before correction, such as: (x ₁ , y ₁ ), (x ₂ , y ₂ ), (x ₃ , y ₃ ) and Three pairs of coordinate points (x′ ₁ , y′ ₁ ), (x′ ₂ , y′ ₂ )(x′ ₃ , y′ ₃ ) of any three corners of the corrected screen, so that a～ f has 6 affine transformation coefficients in total, and then the current unknown point (X, Y) is brought into formula 5 to obtain the value of the new coordinate point (x', y').

Referring to Figure 13, the process of the interpolation correspondence method in the actual application stage includes the following steps:

Capture a user's eye image 401 (step 611), and use the eye image 401 to locate a pupil center 41 and a reflective point 42 (step 612); use the pupil center and reflective point to define a distance value Li and An angle value θi (step 613); please refer to the description of FIG. 2 for the above steps, and will not repeat them here.

Next, correspond the distance value Li and the angle value θi (that is, X and Y) to obtain a coordinate point in the corrected normalized distribution map, and then convert the coordinate point to the position corresponding to the screen by interpolation (step 614), so That is, the screen positioning function is completed (step 615).

4. Conclusion

As can be seen from the above description, the pupil positioning method and system of the present invention adopt non-contact technology and have the following advantages:

1. There is no need to wear additional equipment on the human body, and the user does not need to spend expensive expenses.

2. The group correspondence method is to use the distance and angle of the pupil center and the reflective point obtained from multiple groups of areas to establish a one-to-one relationship with the space of multiple groups of areas on the screen, so the positioning function can be accurate and accurate without complicated calculations. effective control.

3. The interpolation correspondence method converts a sector drawn by a small number of distance-angle points into a relatively regular rectangle by a twisting operation, and then maps the rectangle to the screen, so that a fairly approximate and good corresponding effect can be obtained, and In the training phase, users only need to watch a few positions to achieve a good positioning effect, which is not only accurate but also saves effort.

Claims (12)

1. pupil positioning method, in order to orient the position that the user watches screen when a user watches a screen attentively, it is characterized in that: described method comprises following step:

(1) acquisition user's eye image, and utilize the eye image to orient a pupil center and a reflective spot;

(2) make a distance value with pupil center to the distance bound of reflective spot, and define a datum line with pupil center location and cooperate reflective spot to define an angle value with respect to a vector of pupil center; And

(3) distance value of diverse location and one angle-range distribution figure that angle value obtains are watched in utilization; And

(4) obtain a coordinate points according to a predetermined way in angle-range distribution figure correspondence with the distance value of watching arbitrary position and angle value, when obtaining whereby watching the screen any point corresponding to the position of screen.

2. pupil positioning method as claimed in claim 1 is characterized in that: described step (3) is that screen is divided the zones of organizing in advance, and each shows wherein one group of zone, and execution comprises the steps:

(a1) show when one group of zone is watched for the user execution in step (1) and step (2) at every turn; And

(a2) if judge that All Ranges has shown when finishing, obtain the distance-angular distribution figure in corresponding each group zone.

3. pupil positioning method as claimed in claim 2 is characterized in that: the predetermined way of described step (4) is a grouping correspondent method, and described grouping correspondent method comprises the steps:

When (a3) user watches the arbitrary position of screen, execution in step (1) and step (2) a distance value and an angle value to obtain the position;

(a4) in angle-range distribution figure, obtain a coordinate points with distance value and angle value; And

(a5) utilize a minimum target letter formula to find out the nearest zone of range coordinate point, borrow described zone to convert corresponding to the position of screen.

4. pupil positioning method as claimed in claim 3 is characterized in that: the formula of described minimum target letter formula is as follows: Min Obj=W1|Dist_t-Dis_i|+W2|Ang1e_t-Angle_i|; Dist_i and Angle_i are target range value and angle value, and Dis_t and Angle_t are present distance value and angle value, and W1 and W2 are for assigning weight.

5. pupil positioning method as claimed in claim 1 is characterized in that: described step (3) is that screen is divided the zones of organizing in advance, and each shows wherein one group of zone, and execution comprises the steps:

(b1) show when one group of zone is watched for the user execution in step (1) and step (2) at every turn;

(b2) if judge that All Ranges has shown when finishing, obtain one and comprise distance-angular distribution figure that each group is regional;

(b3) will turn round positive computing to obtain a regular distribution plan through one apart from-angular distribution figure; And

(b4) utilize affine switch technology to proofread and correct regular distribution plan.

6. pupil positioning method as claimed in claim 5 is characterized in that: the predetermined way of described step (4) is an interpolation correspondent method, comprises the steps:

When (b5) user watches the arbitrary position of screen, execution in step (1) and step (2) a distance value and an angle value to obtain the position;

(b6) obtain a coordinate points with distance value and angle value correspondence in the regular distribution plan after correction; And

(b7) coordinate points is converted corresponding to the position of screen with interpolation method.

7. pupil positioning system, it is characterized in that: the pupil positioning system comprises:

One light source is launched a light beam irradiates one user's eye;

One video camera, acquisition user's eye image; And

One computing module has:

One feature location unit is obtained the eye image and is oriented a pupil center and a reflective spot,

One distance-angle processing unit is made a distance value with pupil center to the distance bound of reflective spot, and defines a datum line with pupil center location and cooperate reflective spot to define an angle value with respect to a vector of pupil center, and

One coordinate transformation unit, after the distance value of diverse location and one angle-range distribution figure that angle value obtains are watched in utilization, obtain a coordinate points according to a predetermined way in angle-range distribution figure correspondence with the distance value of watching arbitrary position and angle value, when obtaining whereby watching the screen any point corresponding to the position of screen.

8. pupil positioning system as claimed in claim 7 is characterized in that: described computing module also has a training unit, and is regional in order to screen is divided in advance many groups, each wherein one group of zone that only shows, and training unit and execution comprise the steps:

The each demonstration when one group of zone supplies the user to watch, the controlling features positioning unit carries out the location and command range-angle processing unit is obtained the regional distance-angle of each group.

9. pupil positioning system as claimed in claim 8 is characterized in that: the predetermined way that described coordinate transformation unit is carried out is a grouping correspondent method, and it comprises the steps:

When the user watched the arbitrary position of screen, the controlling features positioning unit was carried out the location and command range-angle processing unit is obtained a distance value and an angle value;

In angle-range distribution figure, obtain a coordinate points with distance value and angle value correspondence; And

Utilize a minimum target letter formula to find out the nearest zone of range coordinate point, borrow described zone to convert corresponding to the position of screen.

10. pupil positioning system as claimed in claim 9 is characterized in that: the formula of described minimum target letter formula is as follows:

Min Obj=W1|Dist_t-Dis_i|+W2|Angle_t-Angle_i|; Dist_i and Angle_i are target range value and angle value, and Dis_t and Angle_t are present distance value and angle value, and W1 and W2 are for assigning weight.

11. pupil positioning system as claimed in claim 7 is characterized in that: described computing module also has a training unit, and is regional in order to screen is divided in advance many groups, each wherein one group of zone that only shows, and training unit and execution comprise the steps:

The each demonstration when one group of zone supplies the user to watch, the controlling features positioning unit carries out the location and command range-angle processing unit is obtained the regional distance-angle of each group;

If judge that All Ranges has shown when finishing, obtain one and comprise distance-angular distribution figure that each group is regional; To turn round positive computing to obtain a regular distribution plan through one apart from-angular distribution figure; And

Utilize affine switch technology to proofread and correct regular distribution plan.

12. pupil positioning system as claimed in claim 11 is characterized in that: the predetermined way of described coordinate transformation unit is an interpolation correspondent method, and it comprises the steps:

When the user watched the arbitrary position of screen, the controlling features positioning unit was carried out the location and command range-angle processing unit is obtained a distance value and an angle value;

Obtain a coordinate points with distance value and angle value correspondence in the regular distribution plan after correction; And

Coordinate points is converted corresponding to the position of screen with interpolation method.

Images