US 11,867,641 B2
Image processing techniques for multi-sensor inspection of pipe interiors
Todd Kueny, Tarentum, PA (US); Justin Starr, Baden, PA (US); and Foster J. Salotti, Verona, PA (US)
Assigned to RedZone Robotics, Inc., Warrendale, PA (US)
Filed by RedZone Robotics, Inc., Warrendale, PA (US)
Filed on May 4, 2021, as Appl. No. 17/307,041.
Application 17/307,041 is a continuation of application No. 16/353,034, filed on Mar. 14, 2019, granted, now 11,029,257.
Claims priority of provisional application 62/643,482, filed on Mar. 15, 2018.
Prior Publication US 2021/0255116 A1, Aug. 19, 2021
Int. Cl. G01N 21/954 (2006.01); G01N 21/88 (2006.01); G01S 17/89 (2020.01)
CPC G01N 21/954 (2013.01) [G01N 21/8851 (2013.01); G01N 2021/8887 (2013.01); G01N 2021/9544 (2013.01); G01N 2021/9548 (2013.01); G01S 17/89 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
obtaining, from a multi-sensor pipe inspection robot that traverses through the interior of a pipe, two or more sets of condition assessment data for the interior of the pipe collected during a single pass through the interior of the pipe;
the two or more sets of condition assessment data comprising a first data type obtained using a first sensor type and a second data type obtained using a second sensor type;
identifying, using a first image processing technique applied to the first data type obtained using the first sensor type, a pipe feature;
using the pipe feature identified to identify data of the second sensor type associated with the pipe feature;
selecting, using a processor, a feature specific image processing technique based on the data identified of the second sensor type associated with the pipe feature;
combining, using a processor, two or more image processing techniques, including the first image processing technique and the feature specific image processing technique, to adjust imaging of a pipe feature; and
forming, using the processor, an image of the interior of the pipe using the two or more image processing techniques.