Abstract: A position detection system includes at least two optical units configured to image a space, a memory, and a processing device interfaced to the memory and the optical units. The processing device is configured to access image data from the first and second optical units and use this data to determine at least one of a current first position and a current second position representing touch points on a display. The processing device can define a polygon having at least four sides based the current first and current second positions and can access the memory to store and retrieve the polygon. If the processing device can determine only one of the current first position or the current second position based on the accessed image data, the processing device can use the previously defined polygon to estimate the other position that was not determined using the accessed image data.

Abstract: A position detection system includes at least two optical units configured to image a space, a memory, and a processing device interfaced to the memory and the optical units. The processing device is configured to access image data from the first and second optical units and use this data to determine at least one of a current first position and a current second position representing touch points on a display. The processing device can define a polygon having at least four sides based the current first and current second positions and can access the memory to store and retrieve the polygon. If the processing device can determine only one of the current first position or the current second position based on the accessed image data, the processing device can use the previously defined polygon to estimate the other position that was not determined using the accessed image data.

Abstract: Embodiments include position detection systems that can identify two touch locations mapped to positions proximate a GUI object, such as a boundary. In response to movement of one or both of the two touch locations, the GUI object can be affected, such as moving the boundary to resize a corresponding object and/or to relocate the boundary, or the GUI object can be selected without movement of the touch locations. Embodiments include single touch gestures, such as identifying a rolling, bending, or other movement occurring while a touch location remains substantially the same and interpreting the movement as an input command. Embodiments may utilize one or more optical sensors having sufficient sensitivity to recognize changes in detected light due to variations in object orientation, makeup or posture caused by the rolling, bending, and/or other movement(s).

Abstract: Software heuristics can be applied to determine which two points of a plurality of potential touch points are likely actual touch points based on a potential touch point's location relative to a predefined touch area and/or a characteristic of a hypothetical touch corresponding to the potential touch point. For instance, a software filter may determine if a potential touch point lies outside the touch area based on comparing coordinates of the potential touch point to boundaries of the predefined touch area. As another example, if the size of the hypothetical touch exceeds a threshold and is in a particular position (e.g., near an edge of the touch area), the potential touch point may be identified as a ghost touch point. As another example, a filter may evaluate whether a shape of the hypothetical touch exceeds a threshold for asymmetry; if so, the potential touch point may be identified as a ghost touch point.