1
$\begingroup$

I am supposed to design a pick-and-place robot. I have a prismatic joint to go down to a particular depth to pick an object which is located in 2D using computer vision. What is the recommended way to detect the height at which the object is placed? The object can be at a maximum vertical distance of 15 inches (around 38 cm).

What do you think is easier to implement for sensing this height? Any suitable cheap sensors for this arrangement? Or a stereo camera system?

$\endgroup$
1
$\begingroup$

Based on your actuator and their sensor you can try force based control, relying on admittance filter. You command the arm to go down and when an external force is applied it means you pushed the object top.

Alternatively, if you don't need millimeter accuracy you can use cheap ultra-sound sensors to detect the depth. Or if you need more precision a 1 point laser range finder. Those two option should be cheap and with simple information to handle. Place the sensor between the claws of you gripper if you can.

In general, if you are not confident with vision I would recommend to avoid stereo camera system, as the calibration and exploitation is not that straight forward. In the vision department you can also look for sensor like kinect or intel euclide. They provide depth info readily usable and based on infrared, which might be an issue in sunny places.

$\endgroup$
1
$\begingroup$

One of the experimental methods is to mount the camera on the actuator itself(called eye in the hand). Then use vision algorithms to detect the object of interest using edge detection techniques and use the images for visual odometry.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.