Vision Tracking Update

A lot has changed on the software side since our last update. Using OpenCV libraries and C++, we are now able to identify an object and track it within our field of view.

Using a camera, the AUV captures an image and converts it to a threshold image. We can then instruct the AUV to filter out certain colors and shapes. Once an object is identified, our AUV will find the coordinates of where it is and will move towards it.

This can be used for many aspects of the RoboSub competition, including the first task of the qualifier, which is locating and moving through an inverted U.

Here the program is identifying and tracking an orange bar.

Here the program is identifying and tracking an orange bar.

The white bar is what the computer sees after the color and shape filters are applied.

This is the threshold or binary image. The white bar is what the computer sees after the color and shape filters are applied.

May 18th, 2016

Posted In: Uncategorized

Leave a Reply

Your email address will not be published. Required fields are marked *