Craig C Sullender

US Patent:

8111919, Feb 7, 2012

Filed:

Mar 12, 2008

Appl. No.:

12/046464

Inventors:

Craig Sullender - Austin TX, US

Assignee:

Eyep, Inc. - Austin TX

International Classification:

G06K 9/34

US Classification:

382180

Abstract:

Embodiments disclosed include methods and systems for encoding one or more region features in connected components labeling including associating one or more labels for an object with a memory structure, the memory structure including the one or more region features; storing the one or more region features in the memory structure, the one or more region features processed in raster order to provide a correspondence between one or more region properties and an original location of the object; enabling the memory structure to receive one or more extents of the one or more region properties at an adjustable precision and with an adjustable data rate, the adjustable precision and the adjustable data rate determined as a function of an amount of detail to be stored; and enabling the memory structure to receive one or more extents at an adjustable data rate determined as a function of an amount of detail to be stored independent of pixel data.

US Patent:

8249348, Aug 21, 2012

Filed:

Feb 8, 2008

Appl. No.:

12/028146

Inventors:

Craig Sullender - Austin TX, US

Assignee:

Eyep Inc. - Austin TX

International Classification:

G06K 9/34

US Classification:

382180

Abstract:

Embodiments disclosed include methods and systems for reusing labels for connected component labeling including assigning one or more labels to one or more groups of raw data representing one or more regions by designating one or more data structures as containing information about the one or more regions; connecting the one or more labels determined to be related; choosing a root label for the connected one or more labels, the root label determined by locating an earliest data element from the one or more groups of raw data; altering a label list of the one or more labels, the label list altered by flagging the root label to include a region label index; and overwriting one or more region label indexes according to the root label.

US Patent:

8280167, Oct 2, 2012

Filed:

Feb 4, 2008

Appl. No.:

12/025738

Inventors:

Craig Sullender - Austin TX, US

Assignee:

Eyep, Inc. - Austin TX

International Classification:

G06K 9/34

US Classification:

382180, 382164, 382173, 382179, 382204, 382305

Abstract:

Embodiments disclosed include methods for connected component labeling including labeling groups of raw data as one or more regions, the labeling including designating one or more data structures as containing information about the one or more regions; designating one or more of the regions as one or more subregions to expose a spatial distribution of one or more region features; and arranging at least one memory array with a 1:1 correspondence to a data array associated with the raw data to enable one or more data structures to include feature labels of the one or more subregions, the 1:1 correspondence enabling acquisition of the one or more region features with a controllable precision.

US Patent:

8340421, Dec 25, 2012

Filed:

Mar 30, 2008

Appl. No.:

12/058706

Inventors:

Craig Sullender - Austin TX, US

Assignee:

EYEP Inc. - Austin TX

International Classification:

G06K 9/34

US Classification:

382173, 382180

Abstract:

Embodiments disclosed include methods and systems for three dimensional connected component labeling, including determining a location value for each of one or more labels, each location value identifying a maximum “y” extent and a maximum “z” extent of an associated label region; storing each of the one or more labels that refer to areas subsumed in a determination of the maximum y extent in the maximum “z” extent as a yzMax location value; buffering in a frame buffer the one or more of labels; and providing access via a three-dimensional kernel to one or more values in a current line buffer and/or a current array and/or a previous array.

US Patent:

8577143, Nov 5, 2013

Filed:

Jul 18, 2012

Appl. No.:

13/552597

Inventors:

Craig Sullender - Austin TX, US

Assignee:

eyeP, Inc. - Austin TX

International Classification:

G06K 9/34

US Classification:

382180

Abstract:

Embodiments disclosed include methods and systems for reusing labels for connected component labeling including assigning one or more labels to one or more groups of raw data representing one or more regions by designating one or more data structures as containing information about the one or more regions; connecting the one or more labels determined to be related; choosing a root label for the connected one or more labels, the root label determined by locating an earliest data element from the one or more groups of raw data; altering a label list of the one or more labels, the label list altered by flagging the root label to include a region label index; and overwriting one or more region label indexes according to the root label.

US Patent:

8577144, Nov 5, 2013

Filed:

Aug 27, 2012

Appl. No.:

13/595979

Inventors:

Craig Sullender - Austin TX, US

Assignee:

eyeP, Inc. - Austin TX

International Classification:

G06K 9/34

US Classification:

382180, 382164, 382171, 382173, 382175, 382176, 70790, 70812

Abstract:

Embodiments disclosed include methods for connected component labeling including labeling groups of raw data as one or more regions, the labeling including designating one or more data structures as containing information about the one or more regions; designating one or more of the regions as one or more subregions to expose a spatial distribution of one or more region features; and arranging at least one memory array with a 1:1 correspondence to a data array associated with the raw data to enable one or more data structures to include feature labels of the one or more subregions, the 1:1 correspondence enabling acquisition of the one or more region features with a controllable precision.

US Patent:

20090196504, Aug 6, 2009

Filed:

Feb 12, 2008

Appl. No.:

12/030003

Inventors:

Craig Sullender - Austin TX, US

International Classification:

G06K 9/34

US Classification:

382180

Abstract:

Embodiments disclosed include methods and systems for assigning one or more labels to one or more segments of data received in an incoming segment to a line buffer for propagated component labeling, including preventing repeated labels in each line of the line buffer by assigning a different label for each of the one or more segments of data received in each line; labeling the incoming segment of the one or more segments of data by adopting a label of an overlapping segment on a prior received line when the overlapping segment does not overlap any other segment of data; labeling the incoming segment of the one or more segments of data by adopting a label of an overlapping segment on a prior received line when the overlapping segment overlaps more than one segment on the incoming segment when the segment is a first segment in the line buffer; and labeling the incoming segment of the one or more segments of data by adopting a label of a last overlapping segment when more than one segment overlaps the incoming segment.

US Patent:

20090257672, Oct 15, 2009

Filed:

Apr 9, 2008

Appl. No.:

12/099833

Inventors:

Craig Sullender - Austin TX, US

International Classification:

G06K 9/40

US Classification:

382260

Abstract:

Provided is a system and method for processing data and images including, but not limited to separating data into a plurality of data planes; performing noise analysis to determine an average noise amplitude and noise distribution for each data plane via a gradient calculation; applying an edge mask to weaken isolated transients in one or more of the data planes; applying a noise filter using one or more levels of filtering to one or more of the data planes; and performing detail recovery combining a composite data plane and filtered and unfiltered data planes according to the noise analysis.