flyfy / stereo-vision-based-system-for-position-and-orientation-estimation-of-robotic-manipulators Goto Github PK

This project forked from aieml/stereo-vision-based-system-for-position-and-orientation-estimation-of-robotic-manipulators

The objective of this work was to build a robot arm with stereo vision technique to identify a standard object and do the pick and place task. We have successfully completed the estimation of position and the orientation parameters (X, Y, Z) and 𝞡y of the Center of Volume for 4 types of Standard objects, Cubes, Cuboids, Cylinders with negligible and rectifiable errors. Necessary actions were taken to minimize the errors in position and orient estimation. An embedded system was developed with user friendly software interface, and has install the vision system into the Fanuc M10 Robot ARM available in SLIIT Robotic lab. Furthermore a gripper was designed and 3D printed the prototype. It was attached to the end of the robot arm. After testing the gripper the relationship between force and voltage as well as the relationship between linear movement of the screw nut along the power screw and velocity was obtained. Moreover a graphical user interface was developed which is able to train the linear regression model, live predict the coordinates of the objects, and which can check the accuracy of the predicted data as well as which can send predicted data to the gripper and the robot arm .And In addition as a future implementation, the embedded system will be upgraded to device which can be implemented with any kind of customized Robot Arms available in the Industry.