Hybrid Vision System Combines Deep Learning, 3D Imaging, and Intelligent Robotic Planning

Kitov has introduced new options to its CorePlus line of stand-alone vision systems. Combining deep learning, 3D imaging, novel algorithms, and intelligent robotic planning technologies, the new hybrid vision systems can now scan larger part sizes in both offline and inline configurations, providing OEMs with additional flexibility for inspecting products of all types in any production environment.

“Challenging inspection tasks require sophisticated visual inspection solutions, and with customer needs changing and evolving every day, Kitov.ai has released two new hybrid vision systems to meet these requirements,” said Adam Tabor, Chief Operating Officer and Co-Founder. “With the full lineup of hybrid vision systems, OEMs can now inspect virtually any product and quickly add value to their operation.”

Joining the Kitov.ai Core product line, the longer robotic reach of the CorePlus system allows it to scan products up to 400 mm high and 1130 mm in diameter compared to 800 mm in the previous Kitov One model. The CorePlus system with turntable product handling option features 856 mm x 1158 mm system footprint. Designed for integration onto a conveyor line, Kitov Inline scans products up to 1150 mm in diameter with a 1900 mm x 1380 mm system footprint.

All Kitov.ai products offer a choice of optical head: a 25 mm lens with 82 x 68 mm and resolution of 2448 x 2048, or a 50 mm lens with 41 mm x 34 mm field of view and a resolution of 2448 x 2048.

With Kitov.ai’s hybrid deep learning systems, operators set up a new inspection plan with a CAD file or by allowing the system to capture a full 3D scan of a reference part, from which the system generates a 3D digital twin model of the physical product. A CMOS camera with multiple brightfield and darkfield lighting elements in a photometric configuration then captures multiple 2D images.

To simplify robotic motion programming, Kitov.ai’s intelligent robot planner uses the 3D model to determine the best location and pose for inspecting each region of interest. Algorithms automatically develop the robot program path without the need for operator input. The CorePlus algorithms also dynamically manage the four banks of LED lights to maximize the image and determine how many images to capture for each test point based on whether it requires 2D, 3D, or deep learning analysis. After acquisition, Kitov.ai’s deep learning software provides unprecedented accuracy for classifying potential defects discovered by 2D/3D machine vision algorithms.

For more information: www.kitov.ai