Autonomous vehicle object detection project
In an autonomous vehicle object detection project, l used polygonal segmentation to annotate objects like pedestrians, vehicles, and traffic signs in video frames. Key steps including selecting the object, using the polygon tool to trace irregular shapes, and refining pre-labeled data manually. Zooming in on complex areas ensure precise alignment with object boundaries. Quality control involved cross-checking annotations and spot-checking for accuracy. By adjusting polygons frames-by-frame.