# Polygon `pf.annotate.polygon` draws the outline of each object as a line. It shows the shape as clearly as a [mask](/pixelflow/visualization/mask/) and leaves the object itself uncovered. ```python import pixelflow as pf from mayaku import from_pretrained predictor = from_pretrained("mayaku-n-seg") result = predictor("street.jpg") detections = pf.from_mayaku(result, labels=predictor.class_names) detections = detections.filter_by_confidence(0.5) detections.masks_to_segments() image = pf.read_image("street.jpg") image = pf.annotate.polygon(image, detections) pf.save_image("street-outlines.jpg", image) ``` ![Colored lines tracing the edges of the people, motorcycles and an umbrella in a street photo](/pixelflow/assets/annotate-polygon.jpg) `polygon` draws each detection's `segments`, its outline as `[[x, y], ...]` points. Mayaku gives masks, so `masks_to_segments()` traces an outline from each mask first. Some converters fill `segments` directly, such as `from_ultralytics` for segmentation models and `from_easyocr` for text: those need no extra step. Detections without an outline are skipped. See [masks and outlines](/pixelflow/core-concepts/detections/#masks-and-outlines) for how outlines are made. ## Parameters | Parameter | Default | | |-----------|---------|---| | `image` | | RGB image, drawn on in place | | `detections` | | the `Detections` whose outlines to draw | | `thickness` | from the image size | line width in pixels: 4 on a 1920×1280 image | | `colors` | the default palette | list of RGB colors, picked by `class_id` (see [colors](/pixelflow/visualization/overview/#colors-follow-the-class)) |