Object tracking is incredibly useful when combined with object detection particularly when the objects of interest are in motion. Using computer vision we can detect and track only the subjects or classes of objects we want, which helps us make sense of the world we’re seeing.
With that in mind, today we’re going to show how simple/easy it is to get object tracking (with unique identifier per object) working in parallel with object detection on megaAI. The demo video is embedded below:
It’s as simple as indicating that you want the object tracking stream from the command line with the
-s parameter. For example:
python3 test.py -s metaout previewout object_tracker -dd
And that’s it, now you have unique IDs and tracking for any object that megaAI detects.
That’s all for now. Stay tuned for more exciting features and use-case examples!
— The Luxonis Team
An FPGA-based SoM with integrated vision, audio, and motion-sensing capability
An embedded platform for combining Depth and AI, built around Myriad X
An open source stereoscopic camera based on Raspberry Pi