US 12,403,373 B2
Detecting objects proximate to a treadmill
Jody Nguyen, New York, NY (US); Alvin Dominguez, New York, NY (US); Spike Cheng, Taipei (TW); and Tobey Hsu, Taipei (TW)
Assigned to Peloton Interactive, Inc., New York, NY (US)
Filed by Peloton Interactive, Inc., New York, NY (US)
Filed on May 5, 2022, as Appl. No. 17/737,716.
Claims priority of provisional application 63/184,635, filed on May 5, 2021.
Prior Publication US 2022/0355179 A1, Nov. 10, 2022
Int. Cl. A63B 71/00 (2006.01); A63B 22/02 (2006.01); A63B 24/00 (2006.01); G01S 13/62 (2006.01); G01S 13/86 (2006.01)
CPC A63B 71/0054 (2013.01) [A63B 22/02 (2013.01); A63B 24/0087 (2013.01); A63B 2071/0081 (2013.01); G01S 13/62 (2013.01); G01S 13/867 (2013.01)] 9 Claims
OG exemplary drawing
 
8. A treadmill, comprising:
a deck;
a belt having a running surface that moves around the deck; and
a computer vision system configured to:
capture images of objects within an area proximate to the treadmill;
detect an object within the captured images of the objects by creating a bounding box around the object based on contours within the captured images;
identify the detected object based on determining a size and color of the detected object; and
determine the identified object is of concern to operations of the treadmill based on the identification of the detected object.