Eraser for touch displays

Abstract

An optical IR touch sensing apparatus configured to determine, based on output signals of light detectors, a light energy value for each light path across a touch surface, and generate a transmission value for each light path based on the light energy value. A processor is then configured to process the transmission values to determine an object reference point on the touch surface where the light is attenuated or occluded by an object and determine, from a set of light paths unaffected by an object, a set of object boundary light paths comprising one or more light paths passing closest to the object reference point. The characteristics of the object may then be determined in dependence on the set of object boundary light paths.

Description

INCORPORATION BY REFERENCE TO ANY PRIORITY APPLICATIONS

Any and all applications for which a foreign or domestic priority claim is identified in the Application Data Sheet as filed with the present application are hereby incorporated by reference under 37 CFR 1.57.

BACKGROUND OF THE INVENTION

Field of the Invention

The present disclosure relates to techniques for detecting and identifying objects on a touch surface.

Description of the Related Art

To an increasing extent, touch-sensitive panels are being used for providing input data to computers, electronic measurement and test equipment, gaming devices, etc. The panel may be provided with a graphical user interface (GUI) for a user to interact with using e.g. a pointer, stylus or one or more fingers. The GUI may be fixed or dynamic. A fixed GUI may e.g. be in the form of printed matter placed over, under or inside the panel. A dynamic GUI can be provided by a display screen integrated with, or placed underneath, the panel or by an image being projected onto the panel by a projector.

There are numerous known techniques for providing touch sensitivity to the panel, e.g. by using cameras to capture light scattered off the point(s) of touch on the panel, by using cameras to directly observe the objects interacting with the panel, by incorporating resistive wire grids, capacitive sensors, strain gauges, etc. into the panel.

In one category of touch-sensitive panels known as ‘above surface optical touch systems’ and known from e.g. U.S. Pat. No. 4,459,476, a plurality of optical emitters and optical receivers are arranged around the periphery of a touch surface to create a grid of intersecting light paths (otherwise known as detection lines) above the touch surface. Each light path extends between a respective emitter/receiver pair. An object that touches the touch surface will block or attenuate some of the light paths. Based on the identity of the receivers detecting a blocked light path, a processor can determine the location of the intercept between the blocked light paths.

For most touch systems, a user may place a finger onto the surface of a touch panel to register a touch. Alternatively, a stylus may be used. A stylus is typically a pen shaped object with at least one end configured to be pressed against the surface of the touch panel. An example of a stylus according to the prior art is shown in FIG. 2. Use of a stylus 60 may provide improved selection accuracy and pointer precision over a simple finger touch. This can be due to the engineered stylus tip 62 providing a smaller and/or more regular contact surface with the touch panel than is possible with a human finger. Also, muscular control of an entire hand in a pen holding position can be more precise than a single finger for the purposes of pointer control due to lifelong training in the use of pens and pencils.

PCT/SE2016/051229 describes an optical IR touch sensing apparatus configured to determine a position of a touching object on the touch surface and an attenuation value corresponding to the attenuation of the light resulting from the object touching the touch surface. Using these values, the apparatus can differentiate between different types of objects, including multiple stylus tips, fingers, palms. The differentiation between the object types may be determined by a function that takes into account how the attenuation of a touching object varies across the touch surface, compensating for e.g. light field height, detection line density, detection line angular density etc.

For larger objects applied to the touch surface, such as palms and board erasers, it is possible to use an interaction map of the touch surface to determine an approximate shape of the object. For example, where an optical IR touch sensing apparatus is used, an attenuation map may be generated showing an area on the touch surface where the light is highly attenuated. The shape of an attenuated area may then be used to identify the position and shape of the touching object. In FIG. 5, an example attenuation map of a board eraser is shown. In a technique known according to the prior art, a rough shape of the eraser can be determined by identifying all points with an attenuation above a threshold value. An approximate centroid and orientation of the eraser may then be determined using the image moments of the identified points. Such techniques are described in “Image analysis via the general theory of moments” by Michael Reed Teague. Once the centroid and orientation of the board eraser are determined, width and height of the board eraser can be found by determining the extent of the identified pixels in the direction of the orientation angle and the normal of the orientation angle.

A known problem with this technique is that larger objects placed onto an optical IR touch sensing apparatus, such as those described above, may cause severe distortion to the touch signal. Many systems use a large number of detection lines and a high density detection line grid density across the touch surface to ensure high touch resolution. Whilst a relatively small number of detection lines may be interrupted by a finger or stylus, a larger object, such as a palm or board eraser, may disrupt a relatively large number of detection lines and cause artefacts in the processed touch signal. This may result in a distorted and noisy interaction map, making the position, size, and orientation of the larger object impossible to detect accurately and reliably.

Therefore, what is needed is a way of improving the identification of position, size, and orientation of large objects touching an optical touch system that mitigates the above problem.

SUMMARY OF THE INVENTION

It is an objective of the disclosure to at least partly overcome one or more of the above-identified limitations of the prior art.

One or more of these objectives, as well as further objectives that may appear from the description below, are at least partly achieved by means of a method for data processing, a computer readable medium, devices for data processing, and a touch-sensing apparatus according to the independent claims, embodiments thereof being defined by the dependent claims.

A first embodiment provides A touch sensing apparatus, comprising: a touch surface, a plurality of emitters, arranged around a periphery of the touch surface, configured to emit a plurality of light beams such that one or more objects touching the touch surface cause an attenuation of at least one of the plurality of light beams; a plurality of detectors, arranged around the periphery of the touch surface, configured to receive light from the plurality of emitters on a plurality of light paths, wherein each detector in the plurality of detectors is arranged to receive light from more than one emitter of the plurality of emitters; and a hardware processor configured to: determine, based on output signals from the plurality of detectors, a plurality of transmission values, each of the plurality of transmission values corresponding to each of the plurality of light paths; determine an object reference point on the touch surface where the light is attenuated or occluded by an object based on the plurality of transmission values; determine one or more unaffected light paths from the plurality of light paths that are not affected by the object based on the plurality of transmission values; determine, from the one or more unaffected light paths that are not affected by the object, one or more object boundary light paths in relation with the object reference point; determine one or more characteristics of the object based on the one or more object boundary light paths.

A method of determining characteristics of an object in contact with a touch surface of a touch sensing apparatus, said touch sensing apparatus comprising: a touch surface, a plurality of emitters arranged around the periphery of the touch surface to emit beams of light such that one or more objects touching the touch surface cause an attenuation of the light; and a plurality of light detectors arranged around the periphery of the touch surface to receive light from the plurality of emitters on a plurality of light paths, wherein each light detector is arranged to receive light from more than one emitter; said method comprising the steps of: determining, based on output signals of the light detectors, a transmission value for each light path; processing the transmission values to determine an object reference point on the touch surface where the light is attenuated or occluded by an object, determining a plurality of light paths unaffected by an object in dependence on the transmission values of the respective light paths, determining, from the plurality of light paths unaffected by an object, a plurality of object boundary light paths comprising one or more light paths passing closest to the object reference point, determining one or more characteristics of the object in dependence on the plurality of object boundary light paths.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will now be described in more detail with reference to the accompanying schematic drawings.

FIG. 1 is a top plan view of an optical touch apparatus.

FIG. 2 shows a cross-section of an IR optical touch apparatus according to the prior art.

FIG. 3 shows a cross-section of an IR optical touch apparatus with a board eraser object applied to the touch surface.

FIG. 4 shows a top-down view of an IR optical touch apparatus with a board eraser object applied to the touch surface.

FIG. 5 shows an example interaction map of a board eraser object applied to a touch surface.

FIG. 6 is a flow chart showing a process for determining characteristics of a large touching object.

FIGS. 7a-7f shows a sequence of steps for determining characteristics of a large touching object.

FIG. 8 shows all detection lines used to determine a rectangular touching object and its centre position overlaid on an interaction map.

FIG. 9 shows a touch system with all the detection lines used to determine a rectangular touching object.

FIG. 10 is plot of object boundary lines values of phi against distance to the object centre.

FIG. 11 is plot of object boundary lines values of phi against distance to the object centre where the object is aligned with the phi axis.

FIG. 12 is plot of object boundary lines values of phi against distance to the object centre showing a noise value to be filtered.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

The present disclosure relates to optical touch panels and the use of techniques for providing touch sensitivity to a display apparatus. Throughout the description the same reference numerals are used to identify corresponding elements.

In addition to having its ordinary meaning, the following terms can also mean:

A “touch object” or “touching object” is a physical object that touches, or is brought in sufficient proximity to, a touch surface so as to be detected by one or more sensors in the touch system. The physical object may be animate or inanimate.

An “interaction” occurs when the touch object affects a parameter measured by the sensor.

A “touch” denotes a point of interaction as seen in the interaction pattern.

A “light field” is the light flowing between an emitter and a corresponding detector. Although an emitter may generate a large amount of light in many directions, only the light measured by a detector from an emitter defines the light field for the emitter and detector.

FIG. 1 is a top plan view of an optical touch apparatus which may correspond to the IR optical touch apparatus of FIG. 2. Emitters 30a are distributed around the periphery of touch surface 20, to project light across the touch surface 20 of touch panel 10. Detectors 30b are distributed around the periphery of touch surface 20, to receive part of the propagating light. The light from each of emitters 30a will thereby propagate to a number of different detectors 30b on a plurality of light paths 50.

FIG. 2 shows a cross-section of an IR optical touch apparatus according to the prior art. In the example apparatus shown in FIG. 2, object 60 will attenuate light propagating along at least one light path 50. In the example shown of FIG. 2, object 60 may even fully occlude the light on at least one light path 50.

Light paths 50 may conceptually be represented as “detection lines” that extend across the touch surface 20 to the periphery of touch surface 20 between pairs of emitters 30a and detectors 30b, as shown in FIG. 1. Thus, the detection lines 50 correspond to a projection of the light paths 50 onto the touch surface 20. Thereby, the emitters 30a and detectors 30b collectively define a grid of detection lines 50 (“detection grid”) on the touch surface 20, as seen in the top plan view of FIG. 1. The spacing of intersections in the detection grid defines the spatial resolution of the touch-sensitive apparatus 100, i.e. the smallest object that can be detected on the touch surface 20. The width of the detection line is a function of the width of the emitters and corresponding detectors. A wide detector detecting light from a wide emitter provides a wide detection line with a broader surface coverage, minimising the space in between detection lines which provide no touch coverage. A disadvantage of broad detection lines may be the reduced touch precision, worse point separation, and lower signal to noise ratio.

As used herein, the emitters 30a may be any type of device capable of emitting radiation in a desired wavelength range, for example a diode laser, a VCSEL (vertical-cavity surface-emitting laser), an LED (light-emitting diode), an incandescent lamp, a halogen lamp, etc. The emitters 30a may also be formed by the end of an optical fibre. The emitters 30a may generate light in any wavelength range. The following examples presume that the light is generated in the infrared (IR), i.e. at wavelengths above about 750 nm. Analogously, the detectors 30b may be any device capable of converting light (in the same wavelength range) into an electrical signal, such as a photo-detector, a CCD device, a CMOS device, etc.

The detectors 30b collectively provide an output signal, which is received and sampled by a signal processor 140. The output signal contains a number of sub-signals, also denoted “transmission values”, each representing the energy of light received by one of light detectors 30b from one of light emitters 30a. Depending on implementation, the signal processor 140 may need to process the output signal for separation of the individual transmission values. The transmission values represent the received energy, intensity or power of light received by the detectors 30b on the individual detection lines 50. Whenever an object touches a detection line 50, the received energy on this detection line is decreased or “attenuated”. Where an object blocks the entire width of the detection line of an above-surface system, the detection line will be fully attenuated or occluded.

In an embodiment, the touch apparatus is arranged according to FIG. 2. A light emitted by emitters 30a is transmitted through transmissive panel 10 in a manner that does not cause the light to TR within transmissive panel 10. Instead, the light exits transmissive panel 10 through touch surface 20 and is reflected by reflector surface 80 of edge reflector 70 to travel along a path 50 in a plane parallel with touch surface 20. The light will then continue until deflected by reflector surface 80 of the edge reflector 70 at an opposing or adjacent edge of the transmissive panel 10, wherein the light will be deflected back down through transmissive panel 10 and onto detectors 30b. An object 60 (optionally having object tip 62) touching surface 20 will occlude light paths 50 that intersect with the location of the object on the surface resulting in an attenuated light signal received at detector 30b.

In an embodiment, the top edge of reflector surface 80 is 2 mm above touch surface 20. This results in a light field 90 which is 2 mm deep. A 2 mm deep field is advantageous for this embodiment as it minimizes the distance that the object needs to travel into the light field to reach the touch surface and to maximally attenuate the light. The smaller the distance, the shorter time between the object entering the light field and contacting the surface. This is particularly advantageous for differentiating between large objects entering the light field slowly and small objects entering the light field quickly. A large object entering the light field will initially cause a similar attenuation as a smaller object fully extended into the light field. The shorter distance for the objects to travel, the fewer frames are required before a representative attenuation signal for each object can be observed. This effect is particularly apparent when the light field is between 0.5 mm and 2 mm deep.

In an alternative embodiment, the transmitted light illuminates a touch surface 20 from within the panel 10. The panel 10 is made of solid material in one or more layers and may have any shape. The panel 10 defines an internal radiation propagation channel, in which light propagates by internal reflections. The propagation channel is defined between the boundary surfaces of the panel 10, where the top surface allows the propagating light to interact with touching objects 7 and thereby defines the touch surface 20. This is achieved by injecting the light into the panel 10 such that the light is reflected by total internal reflection (TIR) in the touch surface 20 as it propagates through the panel 10. The light may be reflected by TIR in the bottom surface or against a reflective coating thereon. In this embodiment, an object 7 may be brought in contact with the touch surface 20 to interact with the propagating light at the point of touch. In this interaction, part of the light may be scattered by the object 7, part of the light may be absorbed by the object 7, and part of the light may continue to propagate in its original direction across the panel 10. Thus, the touching object 7 causes a local frustration of the total internal reflection, which leads to a decrease in the energy (or, equivalently, power or intensity) of the transmitted light.

The signal processor 140 may be configured to process the transmission values so as to determine a property of the touching objects, such as a position (e.g. in a x,y coordinate system), a shape, or an area. This determination may involve a straight-forward triangulation based on the attenuated detection lines, e.g. as disclosed in U.S. Pat. No. 7,432,893 and WO2010/015408, or a more advanced processing to recreate a distribution of attenuation values (for simplicity, referred to as an “attenuation pattern”) across the touch surface 20, where each attenuation value represents a local degree of light attenuation. The attenuation pattern may be further processed by the signal processor 140 or by a separate device (not shown) for determination of a position, shape or area of touching objects. The attenuation pattern may be generated e.g. by any available algorithm for image reconstruction based on transmission values, including tomographic reconstruction methods such as Filtered Back Projection, FFT-based algorithms, ART (Algebraic Reconstruction Technique), SART (Simultaneous Algebraic Reconstruction Technique), etc. Alternatively, the attenuation pattern may be generated by adapting one or more basis functions and/or by statistical methods such as Bayesian inversion. Examples of such reconstruction functions designed for use in touch determination are found in WO2009/077962, WO2011/049511, WO2011/139213, WO2012/050510, and WO2013/062471, all of which are incorporated herein by reference.

For the purposes of brevity, the term ‘signal processor’ is used throughout to describe one or more processing components for performing the various stages of processing required between receiving the signal from the detectors through to outputting a determination of touch including touch co-ordinates, touch properties, etc. Although the processing stages of the present disclosure may be carried out on a single processing unit (with a corresponding memory unit), the disclosure is also intended to cover multiple processing units and even remotely located processing units. In an embodiment, the signal processor 140 can include one or more hardware processors 130 and a memory 120. The hardware processors can include, for example, one or more computer processing units. The hardware processor can also include microcontrollers and/or application specific circuitry such as ASICs and FPGAs. The flowcharts and functions discussed herein can be implemented as programming instructions stored, for example, in the memory 120 or a memory of the one or more hardware processors. The programming instructions can be implemented in machine code, C, C++, JAVA, or any other suitable programming languages. The signal processor 140 can execute the programming instructions and accordingly execute the flowcharts and functions discussed herein.

FIG. 4 shows an example of a rectangular board eraser object 310 as applied to the touch surface. Object 310 has a width 430 and a height 420. The orientation angle of object 310 is determined to be the angle 415 between wide axis 405 and touch surface x-axis 400. As object 310 is a rectangle, narrow axis 410 is perpendicular to wide axis 405.

FIG. 5 shows an attenuation map of the board eraser of FIG. 4.

FIG. 6 shows a flow diagram according to an embodiment.

In step 610 of FIG. 6, the signal processor 140 receives and samples output signals from detectors 30b.

In step 620, the output signals are processed for determination of the transmission values (or ‘transmission signals’). As described above, the transmission values represent the received energy, intensity or power of light received by the detectors 30b on the individual detection lines 50.

In step 630, the signal processor 140 is configured to process the transmission values to determine the presence of one or more touching objects on the touch surface. In an embodiment, the signal processor 140 is configured to process the transmission values to generate a two-dimensional attenuation map of the attenuation field across the touch surface, i.e. a spatial distribution of attenuation values, in which each touching object typically appears as a region of changed attenuation. From the attenuation map, two-dimensional touch data may be extracted and one or more touch locations may be identified. The transmission values may be processed according to a tomographic reconstruction algorithm to generate the two-dimensional attenuation map of the attenuation field.

In one embodiment, the signal processor 140 may be configured to generate an attenuation map for the entire touch surface. In an alternative embodiment, the signal processor 140 may be configured to generate an attenuation map for a sub-section of the touch surface, the sub-section being selected according to one or more criteria determined during processing of the transmission values.

In an alternative embodiment, the signal processor 140 is configured to process the transmission values to determine the presence of one or more touching objects on the touch surface by determining intersections between attenuated or occluded detection lines, i.e. by triangulation. In yet another embodiment, the signal processor 140 is configured to process the transmission values to determine the presence of one or more touching objects on the touch surface using non-linear touch detection techniques such as those described in US patent application publication 20150130769 or 20150138105.

In step 640, the signal processor 140 is configured to identify large touching objects having an interaction area greater than a minimum size. In an embodiment, an area of the attenuation map defined by the object interaction is identified. Within the identified area, the point having the highest attenuation value is identified. Then, a flood fill algorithm is applied, starting from the highest attenuation point and selecting all contiguous points in the attenuation map having an attenuation value within a value range of the highest attenuation point. If a sufficiently large number of pixels are selected by the flood fill, the object is determined to be a large object. Alternatively, if the largest diameter of the shape defined by the flood fill is greater than a threshold, the object is designated as a large object. For example, if eraser objects of at least 30 mm in width are determined to be large objects and the reconstructed pixels (i.e. co-ordinates in the attenuation map) are 5 mm in average diameter, the diameter threshold for designating a large object is set to 6 pixels. In one embodiment, an initial size estimate for an object may be determined using centroid and image moments. If the initial size estimate is sufficiently large enough a more detailed analysis using, for example, a flood fill analysis may be used to accurately determine size.

In some embodiments, all touching objects are processed according to the following steps and not just large objects.

In step 650, for each identified large touching object, the signal processor 140 is configured to determine an object reference point 110 within the interaction area of large touching object 310. In one embodiment, an image moment is applied to the attenuation map (e.g. that shown in FIG. 5) to determine a centroid of a detected touching object, for use as the object reference point. E.g. For a scalar attenuation map with pixel intensities I(x,y), raw image moments M_ij are calculated by:

$M_{ij}=\sum _x\sum _y{x}^i{y}^{i}I\left(x,y\right)$

The centroid of the image moment may be calculated as:{°x,°y°}={M₁₀/M₀₀,° M₀₁/M₀₀°}

In another embodiment, signal processor 140 is configured to determine an object reference point within the interaction area of large touching object by determining a local maxima (i.e. point of highest attenuation) in the area of the attenuation map covered by the object. In another embodiment, signal processor 140 is configured to determine an object reference point within the interaction area of large touching object by selecting a point at random within the boundary of the large touching object.

In step 660, the signal processor 140 is configured to determine a set of relatively un-attenuated or un-occluded detection lines passing closest to the object reference point. In an embodiment, the detection lines from each emitter are analysed to determine the detection line passing closest to the object reference point. Then, the first detection lines clock-wise and counterclockwise relative to the closest detection line that are determined to be unaffected (i.e. not substantially attenuated or occluded by the object) are determined to be boundary detection lines, partially defining the boundaries of the object. This process is repeated for each emitter until boundary detection lines for the object are determined for each emitter.

In more detail, an embodiment of step 660 includes the following steps:

1) The emitter/detector pairs forming each detection line are analysed in a counter-clock-wise direction. As shown in FIG. 7a, the detection line from the first emitter e0 on the bottom side of the touch surface and the first detector d0 on the right side is the first detection line to be analysed. For the purposes of clear explanation, the touch system shown in FIG. 7 shows only emitters along left and bottom edges and detectors along the right and top edges. However, it is understood that the present concepts may be applied to touch systems having a variety of emitter and detector geometries including interleaved emitter and detector arrangements.

2) As shown in FIG. 7b, the detector counter is then incremented in clock-wise direction (i.e. d_i+1) and the detection line between emitter e₀ and the incremented detector d_i+1 is analysed. This loop continues and the detection lines from the emitter are therefore analysed in a counterclockwise pattern until the intersect detection line 150 from e₀ to d_k and passing through the object reference point 110, or determined to pass closest to the object reference point 110, is found.

3) FIG. 7c shows the boundary lines 160 and 170. Boundary lines 160 and 170 are determined to be the detection lines that:

• • Originate from e₀
  • Have the closest angular distance to intersect detection line 150 in a clock-wise and counterclockwise direction,
  • Are determined to be unaffected (i.e. not substantially attenuated or occluded) by object 310.

In FIG. 7c, boundary lines 160 and 170 are shown to be detection lines between e0 and d_cw,0 and e₀ and d_ccw,0 respectively (where subscript cw or ccw indicate the direction and 0 indicates the corresponding emitter number). In an embodiment, a detection line is determined to be unaffected by an object when it is attenuated (or occluded) by less than 50%. E.g. The transmission value of the detection line is greater than 50% of its maximum value when there is no touch/pen/eraser interaction. Preferably, the maximum transmission value for a detection line is determined in dependence on an estimated background level for each detection line. An estimated background level may be maintained and updated only when there are no interactions with the touch surface. Alternatively, a background level can be measured during a calibration step during manufacturing.

For FTIR based systems, the attenuation resulting from object interaction with the touch surface is substantially lower. Therefore, the differentiation between an affected and unaffected detection line is dependent on a correspondingly lower attenuation threshold (e.g. 0.1%-10%).

In an embodiment, the first unaffected detection line identified in a clockwise search from intersect detection line 150 is set to be boundary line 160. Similarly, the first unaffected detection line identified in a counterclockwise search from intersect detection line 150 is set to be boundary line 170

4) As shown in FIG. 7d-f, after boundary lines 160 and 170 for object 310 have been determined for e₀, the next emitter in the counterclockwise direction (e.g. e_i+1) is used and steps 1-3 are repeated. This process is repeated until boundary lines 160 and 170 for object 310 are determined for every emitter.

As the emitter/detectors are processed in a circular (and clock-wise) order, a geometric consequence is that the detection line defined by [e_j+1,d_k] will be further away (clock-wise direction) from the object reference point 110 than [e_j,d_k]. Therefore, in a preferable configuration, when detection lines for the next emitter in the clock-wise direction are analysed, the first detection line to be analysed may be [e_j+1,d_k] and then continued in a clock-wise direction. This allows a significant reduction in the number of computations required to determine the set of object boundary lines.

In step 670 of FIG. 6, the signal processor 140 is configured to determine characteristics of the large touching object in dependence on the set of uninterrupted detection lines passing closest to the point.

1) A true centre point 140 of a rectangle object 310 (as opposed to object reference point 110) can now be found as the solution to the following over-determined set of linear equations, solved using normal equations.

A normal vector (having unit length) for each emitter-to-detector detection line as well as a position on each detection line (which can be the geometrical position of either emitter or detector or some other point) is determined.

For all emitters j, we get two linear equations for each emitter:dot_product(normal(e_j−d_cw,j)center−position_j)=0dot_product(normal(e_j−d_ccw,j)center−position_j)=0

Where normal is the normal vector and position_j is a position along the detection line. Then, all of the linear equations are solved to determine a centre position 140.

This technique also allows a centre position to be determined for regular shapes, oblongs, etc.

2) Once centre point 140 has been determined, all determined boundary lines for all emitters are analysed to determine their angle φ (phi), defined as the angle between the normal to the detection line and the touch surface x-axis 400, and the shortest distance from centre point 140 to the detection line. FIG. 10 shows a typical plot of all boundary line values of phi against distance to centre point 140. FIG. 10 shows both positive and negative distances as the distance is the projection onto the normal. Negative values correspond to detection lines whose normal (defined by the normal to the direction from detector to emitter) points away from the point. Boundary lines defined by e_i and d_cw,j will have the opposite sign to boundary lines defined by e_j and in FIG. 10.

3) The boundary line with the smallest magnitude distance from centre point 140 is then determined (defined here as the ‘selected boundary line’). The characteristics of the selected boundary line will provide information useful information about the characteristics of object 310. First, where the object is substantially rectangular, e.g. as shown in FIG. 4, the width 430 of object 310 the may be determined in dependence on the vector 200 defining the shortest distance from the selected boundary line to centre point 140. As the object is rectangular, the magnitude of the vector may be assumed to be half of width 430. Therefore width 430 of object 310 may be determined to be twice the magnitude of the vector.

Furthermore, the angle of the vector 200 also defines the orientation angle 440 of rectangular object 310. The angle phi of vector 200 defines the narrow axis 410 of object 200. Consequently, the angle of the wide axis 405 of the rectangle may be defined as

$phi\pm \frac{\pi }{2}\mathrm{in}\mathrm{phi}\pm \frac{\pi }{2}.$ Using

$\mathrm{phi}\pm \frac{\pi }{2},$ we can also use vector 200 of the boundary line located at FIG. 10 to determine the height 420 of the object. Similar to above, height 420 of object 310 may be determined to be twice the magnitude of the vector 200 of the boundary line located at

$phi\pm \frac{\pi }{2}.$

In one embodiment, the phi and height values for object 310 are determined using an average of a plurality of the smallest values.

FIG. 8 shows an attenuation map of an object 310 with an overlay of all determined boundary lines. FIG. 9 shows an equivalent set of boundary lines without the underlying attenuation map.

FIG. 11 shows an example plot of object boundary lines values of phi against distance to the object centre, similar to FIG. 10, but wherein the object is located at the axis of phi. The plot effectively wraps at 0 and pi.

In system where detection lines are unstable or unreliable due to ambient noise or other factors, the touch system may invalidate certain detection lines, all detection lines from one emitter, or all detection lines to one detector, from use. In an embodiment, all detection lines determined to be invalid by the system are excluded from use as boundary lines. Similarly, where one or more detection line has been recently re-validated after a period of being invalidated, the one or more detection lines are still excluded from use as boundary lines for a period of time (e.g. between 1 and 500 ms.)

If the system determines that a single boundary line of a pair of boundary lines from an emitter is unstable or unreliable, the pair of boundary lines will not be used for determination of the centre point 140. However, the other boundary line of the pair may still be used for determination of object characteristics.

Other considerations for handling the noise in a touch system are described. In one embodiment, where intersect detection line 150 is determined to be unaffected by object 310 or otherwise too noisy, the boundary lines 160 and 170 for the corresponding emitter are discarded. In another embodiment, boundary lines having an angular distance that is significantly smaller than an average for a small phi-region are analysed and pruned. As shown in FIG. 12, a single boundary lines at phi=2.6 is too close to the centre and hence it is removed from consideration. Similarly, any boundary lines that lie too far from a median or average of nearby boundary lines may also be filtered.

In one embodiment, pairs of boundary lines 160 and 170 are discarded when their mutual angle is larger than a threshold angle. Preferably, the threshold angle is defined as:

$\arctan \left(\frac{\mathrm{maximum}\mathrm{allowable}\mathrm{object}\mathrm{height}}{\mathrm{distance}\mathrm{between}\mathrm{centre}\mathrm{point}140\mathrm{and}\mathrm{emitter}}\right)$ or any similar approximation. This allows exclusion of pairs of boundary lines which would indicate an object above an allowable size. Other suitable threshold angles, such as fixed angles, may also be used. Removal of boundary lines where the mutual angle exceeds a threshold reduces problems with other occluding touches/erasers.

In one embodiment, the shape of the object 310 is optimized to the plotted data of FIG. 10 using many or all of the plotted boundary lines. An example of a suitable optimization algorithm is an algorithm for maximising rectangular area whilst fitting within the boundary lines. This optimization may include the use of boundary lines as defined in the various ways above.

A determination of if the shape of the found rectangle is more similar to a circle than a square/rectangle may be performed by analysis of the distance values between the boundary lines 160 and 170 of each pair on FIG. 10. A rectangular shape may be expected to have a variation of at least sqrt(2)*rectangular width 430 between the boundary values. A circle, on the other hand, would be expected to have a substantially consistent distance between values of boundary lines 160 and 170 pairs in FIG. 10.

In one embodiment, a statistical analysis of the plotted boundary lines is performed to determine the quality of the shape data, e.g. the noisiness of the shape as compared with an expected shape. Where the shape is determined to be noisy and a confidence in the data to match a known shape is determined to be low, the system is configured to fall back to shape and position determined by flood filled attenuation map or other alternative methods of determining shape data.

In an alternative embodiment, boundary lines 160 and 170 are determined to be the detection lines that:

• • Originate from e₀
  • Have the further angular distance to intersect detection line 150 in a clock-wise and counterclockwise direction,
  • Are determined to be affected (i.e. substantially attenuated or occluded) by object 310.

In this embodiment, a detection line is determined to be affected by an object when it is attenuated (or occluded) by more than 50%. E.g. The transmission value of the detection line is lower than 50% of its maximum value when there is no touch/pen/eraser interaction. In this embodiment, boundary lines 160 and 170 may be determined to be the last affected detection when moving clockwise (for boundary line 160) or counterclockwise (for boundary line 170) away from intersect detection line 150.

In this embodiment, the boundary lines are determined to be the set of detection lines having a transmission value below a first threshold and passing within the boundaries of the object but furthest from the object reference point. In an FTIR embodiment of the above, the attenuation resulting from object interaction with the touch surface is substantially lower. Therefore, the differentiation between an affected and unaffected detection line is dependent on a correspondingly lower attenuation threshold (e.g. 0.1%-10%).

In another embodiment, a combination of boundary lines calculated according to the above embodiments may be used in combination to determine an improved accuracy of object characteristics determination.

It is noted that the order in which detection lines are analysed is arbitrary. It is possible to start with fixed emitters or detectors when search for intersect detection lines and boundary lines

Claims

1. A touch sensing apparatus, comprising: a touch surface,a plurality of emitters, arranged around a periphery of the touch surface, configured to emit a plurality of light beams such that one or more objects touching the touch surface cause an attenuation of at least one of the plurality of light beams;a plurality of detectors, arranged around the periphery of the touch surface, configured to receive light from the plurality of emitters on a plurality of light paths, wherein each detector in the plurality of detectors is arranged to receive light from more than one emitter of the plurality of emitters; anda hardware processor configured to: determine, based on output signals from the plurality of detectors, a plurality of transmission values, each of the plurality of transmission values corresponding to each of the plurality of light paths;determine an object reference point on the touch surface where the light is attenuated or occluded by an object based on the plurality of transmission values;determine, from the plurality of light paths, one or more object boundary light paths passing close to a surface of the object based on the plurality of transmission values;determine one or more characteristics of the object based on the one or more object boundary light paths,wherein the one or more object boundary light paths comprises a closest boundary light path passing closest to the object reference point.

2. The touch sensing apparatus of claim 1, wherein the one or more object boundary light paths further comprises a closest perpendicular boundary light path, the closest perpendicular boundary light path being the object boundary light path passing closest to the object reference point and having an angle

3. The touch sensing apparatus of claim 1, wherein determining characteristics of the object comprises determining at least one of the centre point, orientation angle, width, and height of the object.

4. The touch sensing apparatus of claim 1, wherein the step of determining characteristics of the object comprises determining an orientation of the object in dependence on the angle of the closest boundary light path.

5. The touch sensing apparatus of claim 1, wherein the step of determining characteristics of the object comprises determining an orientation of the object in dependence on the angle of the closest boundary light path and the closest perpendicular boundary light path.

6. The touch sensing apparatus of claim 1, wherein the step of determining characteristics of the object comprises determining a limit of a first surface of the object in dependence on the closest boundary light path.

7. The touch sensing apparatus of claim 1, wherein the step of determining characteristics of the object comprises determining a position of a first surface of the object in dependence on the closest boundary light path and the closest perpendicular boundary light path.

8. The touch sensing apparatus of claim 1, wherein the object boundary light paths are not affected by the object.

9. The touch sensing apparatus of claim 1, wherein the object boundary light paths are affected by the object.

10. The touch sensing apparatus of claim 1, wherein the object boundary light paths comprise light paths that are affected by the object and light paths that are not affected by the object.

11. The touch sensing apparatus of claim 1, wherein a light path is determined to be affected by an object when it is occluded by more than 50%.

12. The touch sensing apparatus of claim 1, wherein a light path is determined to be affected by an object when the transmission value of the light path is lower than 50% of its maximum value.

13. The touch sensing apparatus of claim 1, wherein a shape of the object is determined in dependence on at least one of the object boundary light paths and on an expected shape.

14. The touch sensing apparatus of claim 13, wherein the shape of the object is matched to the object boundary light paths in dependence on the expected shape.

15. The touch sensing apparatus of claim 13, wherein the expected shape is a rectangle.

16. The touch sensing apparatus of claim 15, wherein the rectangle comprises a set of long edges and a set of short edges, the long edges being at least sqrt(2) longer than the short edges.

17. The touch sensing apparatus of claim 13, wherein the shape of the object is matched to the object boundary light paths in dependence on the expected shape by maximizing rectangular area within the object boundary light paths.

18. The touch sensing apparatus of claim 1, wherein an orientation of the object is determined from the object boundary light paths in dependence on an expected shape, wherein the expected shape is a rectangle.

19. The touch sensing apparatus of claim 1, wherein the object reference point is determined in dependence on a centroid of the object.

20. A method of determining characteristics of an object in contact with a touch surface of a touch sensing apparatus, said touch sensing apparatus comprising: a touch surface,a plurality of emitters arranged around a periphery of the touch surface, configured to emit a plurality of light beams such that one or more objects touching the touch surface cause an attenuation of at least one of the plurality of light beams; anda plurality of detectors, arranged around the periphery of the touch surface, configured to receive light from the plurality of emitters on a plurality of light paths, wherein each light detector in the plurality of detectors is arranged to receive light from more than one emitter of the plurality of emitters;said method comprising the steps of:determining, based on output signals from the plurality of detectors, a plurality of transmission values, each of the plurality of transmission values corresponding to each of the plurality of light paths;determine an object reference point on the touch surface where the light is attenuated or occluded by an object based on the plurality of transmission values;determining, from the plurality of light paths, one or more object boundary light paths passing close to a surface of the object based on the plurality of transmission values;determining one or more characteristics of the object based on the one or more object boundary light paths.