Granted a Major Touch Panel Patent

Apple wins a patent for the longest patent title in the world: "Bottom-up watershed dataflow method and region-specific segmentation based on historic data to identify patches on a touch sensor panel." With that aside, Apple has been granted a patent that relates to touch sensor panel techniques for identifying touch patches.

Apple states that users of touch devices could touch the sensor panels using different fingers, hand positions and orientations, and apply these touches with varying degrees of force. It could be difficult to properly interpret the touch events detected on the sensor panel and take appropriate actions. This is what Apple's patent addresses in detail.

About the Patent Figures

Apple's patent FIG. 1 illustrates an exemplary computing system implementing pre-watershed processing, the watershed algorithm, and post-watershed processing; Patent FIG. 3a illustrates an exemplary thumb and two patches generated as a result of over-segmentation of the thumb touching the touch sensor panel; Patent FIG. 3b illustrates the exemplary thumb and the two patches of FIG. 3a merged into a single patch using spatial smoothing; Patent FIG. 4 illustrates an exemplary one-dimensional image of touch in which the two adjacent peaks separated by a saddle point are compared to establish criteria for merging according to one embodiment of this invention.

Apple's patent FIG. 5 illustrates exemplary patches caused by a right hand touching a sensor panel and the merging of selected patches; Patent FIG. 6 illustrates exemplary bounding boxes that could be established around each patch to limit the number of iterations in post-watershed merging algorithms; And finally patent FIG. 7 illustrates an exemplary image of touch at an edge of a touch sensor panel and a methodology for computing a centroid with enhanced accuracy.

Technically Speaking

Apple's patent relates to applying a watershed algorithm to pixels and their touch values obtained from a scan of a touch sensor panel to determine patches corresponding to images of touch. Prior to applying the watershed algorithm, background pixels having little or no touch values could be eliminated from consideration, and some amount of smoothing could be performed on the pixels to connect overly segmented patches. After the watershed algorithm has been applied, patches whose pixels are substantially linear (e.g. lie within a single row or column), or patches having a very small number of pixels could be eliminated from consideration as being attributed to noise or other errors.

A primary merge algorithm could then be utilized to merge adjacent patches together when the saddle point between them is shallow as compared to the peak represented by the patches. However, if two candidate patches for merging have a total number of pixels below a certain threshold, these two patches may not be merged under the assumption that the patches might have been caused by different fingertips. Conversely, if two candidate patches for merging have a total number of pixels above a certain threshold, these two patches could be merged under the assumption that the patches were caused by a single thumb or palm. Alternatively or in addition, if two candidate patches for merging were part of a single large patch in a previous panel scan, these two patches could be merged under the assumption that the two patches were generated from the same finger or palm.

On Larger Patches like a Face or Ear

Additionally, if a small patch is located substantially directly below a medium-sized patch, this could be indicative of a fingertip and phalange patch, and these two patches could be merged. If a large object such as a face or ear is expected, the merging criteria described above can be made more aggressive under the assumption that the patches were generated from a single large object. Based on the location of the identified patches, hand or finger orientations could be determined, which could be used to further identify which patches could be merged together. Bounding boxes defined around each patch could limit the number of pixels that need to be iteratively considered during post-watershed processing.

Improving Centroid Calculations

Centroids, which define the center point of patches, could be used in post-watershed processing. However, partial patches formed at the edges of the touch sensor panel due to the presence of a finger only partially over the touch sensor panel could lead to inaccurate centroid calculations. To improve the accuracy of centroid calculations for partial patches at the edges of touch sensor panels, a percentage of the pixels in the last column of the patch could be duplicated into an artificial phantom column of pixels that could be used to generate a more accurate centroid location.

Apple credits Wayne Westerman and Joshua Strickon as the inventors of Granted Patent 7,916,126 originally filed in Q2 2007 just as the iPhone was about to debut.

Other Granted Patents of Interest Published Today

• Acoustic noise reduction in power supply inductors
• Methods and apparatuses for restoring color and enhancing electronic images
• Method and apparatus for an intuitive digital image processing system that enhances digital images
• Transient detection by power weighted average (Related to Audio)

Notice: Patently Apple presents only a brief summary of patents with associated graphic(s) for journalistic news purposes as each such patent application and/or Granted Patent is revealed by the U.S. Patent & Trade Office. Readers are cautioned that the full text of any patent application and/or Issued Patent should be read in its entirety for further details. Patents shouldn't be digested as rumors or fast-tracked according to rumor time tables. Apple patents represent true research that could lead to future products and should be understood in that light. About Comments: Patently Apple reserves the right to post, dismiss or edit comments.