Touchscreen devices with improved input accuracy

Abstract

A device improves touchscreen input accuracy by predicting and presenting an intended key before contact and confirming the input only upon touch. A proximity detection module senses approach of at least one input tool, such as a finger or stylus. Using proximity data, a predicted input display module selects one input tool (for example, the tool closest to the touchscreen) and visually presents a corresponding predicted key near a caret, above a keyboard, or beneath the tool, emphasized by magnification, highlighting, or borders. An input confirmation module confirms the selected tool's predicted key as the final input when contact occurs; if multiple tools are present, contact by any tool triggers confirmation of the predicted key corresponding to the previously selected tool. This three-stage flow reduces mis-touches and improves usability across smartphones, tablets, wearables, HMIs, kiosks, and industrial panels, and is compatible with capacitive, infrared, or ultrasonic sensing.

Description

TOUCHSCREEN DEVICES WITH IMPROVED INPUT ACCURACY

[0001] The present invention relates to user-input technologies for electronic devices and, more particularly, to proximity detection and predictive input display techniques that improve input accuracy on devices employing touchscreen-based virtual keyboards.

[0002] Compact displays force virtual keys to be small, increasing mis-touches. Conventional touch systems typically generate input only after physical contact and rely on post-entry correction; they do not provide pre-contact, real-time confirmation of intended input, particularly in multi-finger scenarios.

[0003] Without admitting that any particular reference is prior art, examples of related disclosures include the following.

[0004] KR 10-1469280: distinguishes proximity touch from direct touch on a touchscreen. When a lower-layer GUI object is displayed and an upper-layer object partially overlaps it, a proximity touch on the upper object followed by lateral motion is recognized, and the upper object moves accordingly. The disclosure concerns layered GUI manipulation and neither predicts a specific character key before contact nor resolves ambiguity among multiple simultaneously approaching input tools by selecting the closest tool and confirming only upon contact.

[0005] KR 10-1269375: a touchscreen device with a display unit, a sensing unit, and a control unit. When windows overlap, a touch on an obscured window brings that window to the front. The disclosure does not use pre-contact signals to predict a key or character, does not handle multi-tool proximity with closest-tool selection, and does not gate input confirmation on subsequent contact.

[0006] KR 10-0984230: generates a proximity recognition signal when an external object enters a threshold distance and remains for a specified time, and a proximity movement signal while within the threshold. A control window (e.g., scroll bar or minimap) appears; content scrolls with movements; the display returns when the object exits. Proximity is used for viewport control rather than discrete key prediction with contact-only confirmation.

[0007] KR 10-1064836: a touch-type character input UI with a keypad window and a character display window. Touch or drag highlights associated keys and outputs the corresponding character. The disclosure relies on contact-based feedback and does not perform pre-contact prediction tied to the closest of multiple approaching tools, nor defer final input until after contact for confirming a previously predicted key.

[0008] In sum, none of the above references teaches pre-contact prediction of a single intended key with explicit closest-tool selection in a multi-tool approach and confirmation only upon subsequent contact, as in the present invention.

[0009] To enable users to see (before physical contact) the key the system predicts they intend to press, including when multiple fingers or styli are near the screen, and to confirm the key only upon contact, thereby preventing input errors in advance and reducing mis-touches.

[0010] In one aspect, a device comprises: (i) a proximity detection module configured to detect, when at least one input tool approaches a touchscreen surface, a proximity state; (ii) a predicted input display module configured, based on proximity data, to select one input tool from among detected input tools and to visually present on the touchscreen a predicted input key corresponding to the selected input tool; and (iii) an input confirmation module configured to confirm, upon contact of the selected input tool, the predicted input key as the input. When a plurality of input tools is present, selection may include, for example, selecting the input tool closest to the touchscreen. Sensing may be capacitive, infrared (IR), and / or ultrasonic; data processing may be implemented in hardware and / or software.

[0011] A three-stage flow (proximity detection, predictive display, and contact-gated confirmation) provides pre-contact visual feedback, prevents input errors in advance, suppresses mis-touches, clarifies intent in multi-tool situations, maximizes input accuracy, and enhances user experience across consumer, automotive, medical, and industrial touch interfaces.

[0012] FIG. 1 is a block diagram of a device including a proximity detection module, a predicted input display module, and an input confirmation module.

[0013] FIG. 2 is a conceptual view in which an approaching input tool triggers emphasized visualization of the predicted key (e.g., magnification, highlighting, border) and optional pre-entry character indication; contact confirms input.

[0014] FIG. 3 is a conceptual view in which predicted characters are shown in an input field without on-keyboard enlargement; contact confirms input.

[0015]

[0016] DETAILED DESCRIPTION OF EMBODIMENTS

[0017]

[0018] Definitions

[0019] “Input tool” means any object operable on a touchscreen, such as a finger, stylus, or touch pen.

[0020] “Proximity” means a non-contact approach to the touchscreen surface detectable by sensors (for example, about 1 cm or less depending on sensor capability; values are illustrative and not limiting).

[0021] “Predicted input key” means a key or character estimated from proximity data and, where available, motion cues.

[0022] “Trajectory stability” means stability of an input tool’s movement path as assessed from proximity data.

[0023] “Selected input tool” means the input tool chosen by the system from among detected tools based on proximity data (for example, the closest or highest-confidence tool) and used to determine the predicted input key.

[0024]

[0025] Unless otherwise specified, the terms “include,” “comprise,” and variations thereof are non-limiting. Singular terms may encompass plural forms unless the context dictates otherwise.

[0026]

[0027] Overall Architecture (FIG. 1)

[0028] The device includes a proximity detection module that outputs proximity data (e.g., distance, approach speed, trajectory stability) for one or more input tools; a predicted input display module that selects a single input tool from the proximity data and renders the corresponding predicted key at a user-perceptible location; and an input confirmation module that suppresses input during mere proximity and confirms only the displayed predicted key when contact occurs.

[0029]

[0030] Proximity Detection

[0031] Modalities may include capacitive sensing of electric-field perturbations, IR emission / reflection analysis for distance and direction, and ultrasonic time-of-flight measurements. Hardware processing can include front-end sensing, analog-to-digital conversion, and digital filtering / denoising with feature extraction (e.g., distance, speed, positional stability). Software processing can include multi-sensor fusion, outlier rejection, and tool discrimination / ranking (e.g., selection of the closest or most stable tool) under real-time latency constraints. In multi-tool scenarios, comparative metrics such as minimum distance and stability select one tool; that selection is passed to the display and confirmation modules to avoid ambiguity.

[0032]

[0033] Predicted Input Display (FIGS. 2 and 3)

[0034] Display positions may include near a caret or cursor, above the keyboard, beneath or near the tool, or a user-specified position. Visual emphasis may include magnification, color highlighting, and border outlining to distinguish the predicted key. The predicted key may update continuously with tool movement; optional animations (e.g., smooth scale-in / fade) balance visibility and distraction. User settings may configure position, emphasis type and intensity, and animation timing.

[0035]

[0036] Input Confirmation

[0037] During proximity without contact, no input event is generated. When the selected input tool contacts the touchscreen surface, the module confirms the currently displayed predicted key as the final input. If any tool of a plurality contacts within a brief interval, the system confirms the predicted key corresponding to the previously selected input tool. Debouncing, short hold times, and hysteresis may be used to reduce inadvertent confirmations.

[0038]

[0039] Illustrative Use Cases

[0040] On a smartphone, two fingers at approximately 4 mm and 8 mm approach the screen; the system selects the closer finger, highlights its predicted key, and confirms that key upon contact. On a tablet, three fingers at approximately 3 mm, 5 mm, and 10 mm approach; the system selects the closest finger; its contact confirms its predicted key.

[0041] By way of example and without limitation, a preferred implementation uses capacitive sensors integrated with a touchscreen controller and IR proximity emitters / receivers along a bezel; a proximity threshold of about 10 mm with hysteresis of 2-3 mm; a brief hold-time / debounce of 30-80 ms to stabilize predictions; selection of the closest tool based on instantaneous distance with a stability check derived from proximity data; predicted-key rendering near the caret with magnification and border outlining; and contact-gated confirmation via a standard touch controller. Hardware signal conditioning (ADC and digital filtering) may be performed on a touchscreen controller or DSP, and higher-level fusion and UI rendering on an application processor. Parameter values are illustrative and may be tuned per device class.

[0042] This section describes exemplary modes for carrying out the invention. The examples are illustrative and non-limiting; parameters may be tuned to the device class and sensor capability.

[0043] 1. System overview and pipeline.

[0044] (a) Proximity sensing: one or more proximity modalities (e.g.,

[0045] capacitive, infrared, ultrasonic) sample at a rate sufficient for real-

[0046] time feedback (illustratively 100-300 Hz). A proximity state is

[0047] recognized when sensor data indicate approach within an illustrative

[0048] threshold (e.g., about 10 mm) with optional hysteresis (e.g., 2-3 mm).

[0049] (b) Signal conditioning: front-end analog conditioning is followed by

[0050] ADC and digital filtering / denoising. Features such as instantaneous

[0051] distance, approach speed, and short-term positional stability are

[0052] computed from proximity data.

[0053] (c) Tool identification and selection: when a plurality of tools is

[0054] detected, the controller selects one input tool (the selected input

[0055] tool) using proximity-derived metrics (e.g., minimum distance with a

[0056] stability check and a brief lock-in window to avoid oscillation).

[0057] (d) Predicted key computation: the selected tool’s proximity

[0058] coordinates are mapped to the layout of an on-screen keyboard. A single

[0059] predicted input key is determined and updated as the tool moves.

[0060] (e) Predictive display: the predicted key is rendered at a user-

[0061] perceptible location (e.g., near the caret, above the keyboard, or near

[0062] the tool) with optional emphasis (magnification, highlighting, border).

[0063] (f) Contact-gated confirmation: no input is generated during mere

[0064] proximity. When the selected input tool contacts the touchscreen, the

[0065] system confirms the currently displayed predicted key as the final

[0066] input. If any tool of the plurality contacts within a short interval,

[0067] the system still confirms the key corresponding to the previously

[0068] selected input tool.

[0069] (g) Fallbacks: if the selected tool exits the proximity state or

[0070] prediction confidence drops below a threshold, the predictive display

[0071] is cleared until proximity resumes.

[0072] 2. Hardware / Software partition.

[0073] Sensing and low-latency filtering may run on a touchscreen controller

[0074] or DSP; higher-level fusion, tool selection, and UI rendering may run

[0075] on an application processor. The partition may be varied to meet power

[0076] and latency targets.

[0077] 3. Multi-tool arbitration details.

[0078] Selection may be based on argmin distance subject to a stability

[0079] constraint, with a short lock-in period to prevent rapid alternation

[0080] between tools that are similarly close. If the selected tool is lost

[0081] (e.g., leaves the proximity envelope), selection may be re-run on

[0082] remaining tools.

[0083] 4. UI behavior and timing.

[0084] Illustratively, prediction is updated within 10-30 ms after proximity

[0085] feature updates. Emphasis strength and animation (e.g., smooth scale-

[0086] in / fade) are user-configurable. A debounce / hold time of about 30-80 ms

[0087] prior to confirmation reduces inadvertent activations.

[0088] 5. Calibration and personalization.

[0089] On first use, the device may calibrate baseline proximity levels and

[0090] touch thresholds. Optional user preferences (prediction display

[0091] position, emphasis type, animation) can be stored per app or globally.

[0092] 6. Robustness and safety.

[0093] Hysteresis, debouncing, and outlier rejection improve stability. When

[0094] multiple contacts occur nearly simultaneously, the system resolves to

[0095] the predicted key of the selected input tool to avoid multi-touch mis-

[0096] entries. If the keyboard is hidden or a non-text control is active,

[0097] prediction may be suppressed.

[0098] 7. Variations.

[0099] The invention applies to full keyboards and other on-screen controls

[0100] (e.g., dialers, PIN pads). Proximity can be implemented with any

[0101] suitable sensor stack; additional modalities may be added without

[0102] departing from the scope of the claims.

[0103] The invention is applicable to smartphones, tablets, notebooks, in-vehicle HMIs, public terminals, medical or industrial panels, and wearables; any environment where clarifying intended input before contact improves accuracy, speed, and safety.

[0104] Not applicable. The present application does not contain a sequence listing and does not disclose nucleotide or amino-acid sequences.

Claims

1.A device comprising:a proximity detection module configured to detect, when at least one input tool approaches a touchscreen surface, a proximity state;a predicted input display module configured to select, based on data delivered from the proximity detection module, one input tool from among a plurality of input tools and to visually display, on the touchscreen, a predicted input key corresponding to the selected input tool; andan input confirmation module configured, when the selected input tool contacts the touchscreen surface, to confirm the predicted input key as an input;wherein, when a plurality of input tools is present, one input tool (a “selected input tool”) is selected based on the proximity data, including selecting the input tool that is closest to the touchscreen or otherwise determined from the proximity data; and wherein, upon contact by any one of the plurality of input tools with the touchscreen surface, the input confirmation module confirms, as the final input, the predicted input key corresponding to the previously selected input tool.2.The device of claim 1, wherein the proximity detection module detects the proximity state based on at least one sensing modality selected from capacitive sensing, infrared sensing, and ultrasonic sensing.3.The device of claim 1, wherein the predicted input display module displays the predicted input key at a position selected from near a cursor on the touchscreen, above a keyboard, beneath the input tool, and a user-specified position.4.The device of claim 3, wherein the predicted input key is displayed with at least one visual emphasis selected from magnification, color highlighting, and border outlining.5.The device of claim 1, wherein the proximity detection module processes data using at least one of a hardware signal-processing unit and a software analysis algorithm.

Images