Abstract: A tactile feedback apparatus for a capacitive sensing device is comprised of a dielectric insulator having a surface. A finger receiving recess is defined in the surface. The finger receiving recess is configured for receiving a deflected portion of a finger. The finger receiving recess is comprised of a finger deflecting feature and a tactile feedback feature. The finger deflecting feature is coupled to the surface and structured such that a variety of finger sizes pressed into the finger receiving recess would deflect into a predictable and repeatable shape for affecting the capacitive sensing device in a substantially uniform manner. The tactile feedback element is coupled to the surface and disposed such that the tactile feedback element contacts the deflected portion of the finger to provide tactile feedback to the finger only when the finger is sufficiently deflected into the finger receiving recess to actuate the capacitive sensing device.

Abstract: Scrolling or other user interface functions are provided in response to two-dimensional inputs produced by a touchpad or other input device. The user initially touches the touch sensor in an area near a periphery of the touch sensor and moves his or her finger along the periphery of the device, and then provides a subsequent motion of the finger across the touch sensor in a direction that departs from the periphery. The amount of scrolling produced on the computing system may be determined as a function of a distance moved by the user's finger.

Abstract: Scrolling or other user interface functions are provided in response to two-dimensional inputs produced by a touchpad or other input device. The user initially touches the touch sensor in an area near a periphery of the touch sensor and moves his or her finger along the periphery of the device, and then provides a subsequent motion of the finger across the touch sensor in a direction that departs from the periphery. The amount of scrolling produced on the computing system may be determined as a function of a distance moved by the user's finger.

Abstract: One-dimensional signals suitable for scrolling or other user interface functions are provided in response to two-dimensional inputs produced by a touchpad or other input device. A current chirality of the two-dimensional inputs is determined to have a first sense if the two-dimensional inputs exhibit a primary clockwise trajectory and a second sense if the two-dimensional inputs exhibit a primarily counter-clockwise trajectory. A one-dimensional variable is accumulated using a magnitude that is a function of a distance reflected by the two-dimensional inputs and a sign that is based upon the current chirality. The one-dimensional inputs to the computing system may then be extracted from the accumulated one-dimensional variable as appropriate. The methods may be modal, thereby allowing the same pointing device to be used for both two-dimensional and one-dimensional input.

Abstract: A ballistics subsystem is coupled to a force sensor of an isometric input device. The ballistics subsystem augments control by applying a dual gain transfer function that smoothly transitions from separate gain factors for low and high force inputs. When applied in a cursor device, such as a joystick or force pad, pointer movement is coupled to input force, not input displacement. Traditionally, an array of miniature strain gauges is used to measure the input force. A more recent method optimized for very low cost uses an array of capacitive sensors. Regardless of what physical mechanism is used to measure input force, customized algorithms are generally used to establish the feel and usability of an isometric joystick. The invention comprises several new methods for optimizing the use of an isometric joystick as a cursor-positioning device, altering the transfer function gain to take advantage of asymmetry, detecting selection and deselection, and combined selection and deselection.

Abstract: The method determines how transformations in a piecewise-polynomial image model affect the model's fidelity. Up to third order two-dimensional polynomials are supported. The fidelity measure is the total squared error between model predictions and pixel values. The transformations supported are merging two domains, adding a single pixel or a group of pixels to a domain, removing a single pixel or a group of pixels from a domain, and altering the intensity value of a sample or group of samples from a surface over a domain. Details are disclosed for applying the method to 256 level images up to 640.times.480 pixels in size. Also disclosed is a procedure for efficiently performing coordinate transformations on two-dimensional moment vectors that extends the method to arbitrarily sized images. A disclosed procedure for efficiently converting a domain's moment vector to a least squares polynomial coefficient vector and automatically zeroing unsupported coefficients makes the method useful for freeform domains.