I am using detection.get_bbox() and xmin() etc to extract detection coordinates. These are normalised, but in what range? [-1,1]? I do not have a camera module, so exploring this with the video resource is extremely chaotic.
Hey @Daniel_Manz,
The bounding box coordinates are normalized to [0, 1], not [-1, 1]. They’re basically percentages of the image size (ranging from 0.0 to 1.0), not actual pixel values.
If you need more details on this, definitely check out the dev guide here: hailo-apps-infra/doc/developer_guide/writing_postprocess.md at 6fdfb5bebced58cd07821f357549bf75e3f91f19 · hailo-ai/hailo-apps-infra · GitHub
Hope this helps!