What is Cortex?
CortexRecognition® is ground breaking visual recognition and visual guidance software.
It functions very similar to the human visual process and can learn a large number of objects then recognize and locate any of the learned objects regardless to its presentation in the visual field of the camera.
See RobEye™ Bin Picking Page
- RRI patented algorithms
- Works similar to our brain
- NOT standard machine vision
- ONE Image is all we need
- NO complex multi camera setup
- NO structured lighting (lasers etc.)
X Y Z Rx Ry Rz!
Ignoring Rx and Ry is not acceptable for all applications!
That's simply true. Stop messing with calibration targets!
NO Structured Light
Frequently Asked Questions
1. How fast can your system identify and localize a part?
Depends on the size of the part and the features recognized. Generally speaking, all of our automotive applications are less than 300ms for recognition, localization, and transmission of data to robot. Of course then the robot has to move
2. What size parts can the system work with?
We have applications in production from 1”x 3” clips all the way up to the entire body side panel of a large SUV
3. What is the uncertainty of the extracted position and orientation (pose)
of a part?
Good question... Most of our applications use our patent pending iterative process where we insure the robot moved to where we told it to go, as well as verify our math. This iterative process insures accuracy to the capable resolution of the robot. We have an application in production, where a robot is guided to insert a piston into an engine cylinder to 100µm.
4. What is the cost?
Submit request form
5. Do you provide the identification and pose extraction software with
CortexRecognition® is the patented recognition engine that enables the Robeye™ system. The system comes complete with software, controller, camera, cables, and lights.
6. What other software is needed to run it and can the output be read
No other software is need. Real time has not been achieved yet due to communication lag between the robot and Robeye™
7. Are there certain kinds of parts that the system does not work well
with (e.e., polished metal, glass, etc.)?
System is most robust on parts with many features and details. System is least robust (in 6 degrees of freedom) on small simple shaped pieces and any piece with little to no contrast.
8. Does the system require a CAD model of the parts to be identified and
localized or can it be trained on a new part without a CAD model?
NO CAD is required. Simple one button point & click = taught object.