3D Segmentation for Robotic Singulation

The logistics industry is poised to rival the automotive sector in its use of industrial robotics. One of the main barriers to growth in this area is the need for quality 3D machine vision segmentation that is able to make sense of a scene in the context of the task and to tell the robot where an object is located so that it can separate it from the other objects for induction into the rest of the already highly automated logistic sorting system. Traditional machine vision techniques and pure AI-based approaches are currently being used in attempt to solve this problem, but a clear winner has yet to emerge. Bruce elaborates on the current state of the art for 3D segmentation and on which technology, if any, is looking like the final winner.

About the presenter
Dave Bruce joined FANUC America Corporation (FAC) in 1997 and is currently an engineering manager for a group of engineers that supports FANUC robot integrators and end users via iRVision and third-party machine vision-enabled applications. He received a master's degree in computer science from Oakland University in Auburn Hills, Mich., and a bachelor's degree in electrical engineering from the University of Windsor in Windsor, Ontario, Canada.