Scott James Jordan

US Patent:

20210107165, Apr 15, 2021

Filed:

Oct 10, 2019

Appl. No.:

16/598106

Inventors:

- Cambridge MA, US
Scott Jordan - Cambridge MA, US

Assignee:

Mitsubishi Electric Research Laboratories, Inc. - Cambridge MA

International Classification:

B25J 13/08
B25J 19/02
G01L 1/04
G06T 7/73

Abstract:

A tactile sensor including a cap having a top surface and an undersurface. The undersurface includes pins, each pin has a mark. A portion of the undersurface is attachable to a device. A camera positioned in view of the marks, captures images of the marks placed in motion by elastic deformation of the top surface of the cap. A processor receives the captured images and determines a set of relative positions of the marks in the captured images, by identifying measured image coordinates of locations in images of the captured images. Determine a net force tensor acting on the top surface using a stored machine vision algorithm, by matching the set of relative positions of the marks to a stored set of previously learned relative positions of the marks placed in motion. Control the device via a controller in response to the net force tensor determined in the processor.

US Patent:

20210107166, Apr 15, 2021

Filed:

Oct 10, 2019

Appl. No.:

16/598124

Inventors:

- Cambridge MA, US
Scott Jordan - Cambridge MA, US

Assignee:

Mitsubishi Electric Research Laboratories, Inc. - Cambridge MA

International Classification:

B25J 13/08
G01L 5/00
G01B 11/16
G06T 7/73
G06K 9/62

Abstract:

A tactile sensor including a camera positioned to capture images of marks. An elastically deformable skin including an outer surface having attributes and an undersurface having pins, ridges, or both. Each undersurface pin or ridge includes a mark. A processor detects displacement of the marks in captured images and compares the displaced positions of the marks in the captured images to stored sets of prelearned positions of marks, based on a distance function, to determine a quality of match value for each set of the prelearned positions of marks. A best quality matched prelearned pattern of forces is determined using a user selected function, to calculate a best matching set of the prelearned positions of marks. Identify a pattern of forces acting on the elastically deformable skin based on the determined best matched prelearned pattern of forces.