vis4d.vis.image

Image Visualization.

class BoundingBoxVisualizer(*args, n_colors=50, cat_mapping=None, file_type='png', width=2, canvas=<vis4d.vis.image.canvas.pillow_backend.PillowCanvasBackend object>, viewer=<vis4d.vis.image.viewer.matplotlib_viewer.MatplotlibImageViewer object>, **kwargs)[source]

Bounding box visualizer class.

__init__(*args, n_colors=50, cat_mapping=None, file_type='png', width=2, canvas=<vis4d.vis.image.canvas.pillow_backend.PillowCanvasBackend object>, viewer=<vis4d.vis.image.viewer.matplotlib_viewer.MatplotlibImageViewer object>, **kwargs)[source]

Creates a new Visualizer for Image and Bounding Boxes.

Parameters:

n_colors (int) – How many colors should be used for the internal color map
cat_mapping (dict[str, int]) – Mapping from class names to class ids. Defaults to None.
file_type (str) – Desired file type. Defaults to “png”.
width (int) – Width of the bounding box lines. Defaults to 2.
canvas (CanvasBackend) – Backend that is used to draw on images.
viewer (ImageViewerBackend) – Backend that is used show images.

__repr__()[source]

Return string representation of the visualizer.

Return type:: str

reset()[source]

Reset visualizer.

Return type:: None

process(cur_iter, images, image_names, boxes, scores=None, class_ids=None, track_ids=None, categories=None)[source]

Processes a batch of data.

Parameters:

cur_iter (int) – Current iteration.
images (list[ArrayLike]) – Images to show.
image_names (list[str]) – Image names.
boxes (list[ArrayLikeFloat]) – List of predicted bounding boxes with shape [N, (x1, y1, x2, y2)], where N is the number of boxes.
scores (None | list[ArrayLikeFloat], optional) – List of predicted box scores each of shape [N]. Defaults to None.
class_ids (None | list[ArrayLikeInt], optional) – List of predicted class ids each of shape [N]. Defaults to None.
track_ids (None | list[ArrayLikeInt], optional) – List of predicted track ids each of shape [N]. Defaults to None.
categories (None | list[list[str]], optional) – List of categories for each image. Instead of class ids, the categories will be used to label the boxes. Defaults to None.

Return type:

None

process_single_image(image, image_name, boxes, scores=None, class_ids=None, track_ids=None, categories=None)[source]

Processes a single image entry.

Parameters:

image (ArrayLike) – Image to show.
image_name (str) – Image name.
boxes (ArrayLikeFloat) – Predicted bounding boxes with shape [N, (x1,y1,x2,y2)], where N is the number of boxes.
scores (None | ArrayLikeFloat, optional) – Predicted box scores of shape [N]. Defaults to None.
class_ids (None | ArrayLikeInt, optional) – Predicted class ids of shape [N]. Defaults to None.
track_ids (None | ArrayLikeInt, optional) – Predicted track ids of shape [N]. Defaults to None.
categories (None | list[str], optional) – List of categories for each box. Instead of class ids, the categories will be used to label the boxes. Defaults to None.

Return type:

None

show(cur_iter, blocking=True)[source]

Shows the processed images in a interactive window.

Parameters:

cur_iter (int) – Current iteration.
blocking (bool) – If the visualizer should be blocking i.e. wait for human input for each image. Defaults to True.

Return type:

None

save_to_disk(cur_iter, output_folder)[source]

Saves the visualization to disk.

Writes all processes samples to the output folder naming each image <sample.image_name>.<filetype>.

Parameters:

cur_iter (int) – Current iteration.
output_folder (str) – Folder where the output should be written.

Return type:

None

class SegMaskVisualizer(*args, n_colors=50, class_id_mapping=None, file_type='png', color_palette=None, canvas=<vis4d.vis.image.canvas.pillow_backend.PillowCanvasBackend object>, viewer=<vis4d.vis.image.viewer.matplotlib_viewer.MatplotlibImageViewer object>, **kwargs)[source]

Segmentation mask visualizer class.

__init__(*args, n_colors=50, class_id_mapping=None, file_type='png', color_palette=None, canvas=<vis4d.vis.image.canvas.pillow_backend.PillowCanvasBackend object>, viewer=<vis4d.vis.image.viewer.matplotlib_viewer.MatplotlibImageViewer object>, **kwargs)[source]

Creates a new Visualizer for Image and Bounding Boxes.

Parameters:

n_colors (int) – How many colors should be used for the color map.
class_id_mapping (dict[int, str]) – Mapping from class id to human readable name.
file_type (str) – Desired file type
color_palette (list[tuple[int, int, int]]) – Color palette for each class, in RGB format (0-255). If None, a random color palette with n_colors is generated automatically. Defaults to None.
canvas (CanvasBackend) – Backend that is used to draw on images
viewer (ImageViewerBackend) – Backend that is used show images

reset()[source]

Reset visualizer for new round of evaluation.

Return type:: None

process(cur_iter, images, image_names, masks, class_ids=None)[source]

Processes a batch of data.

Parameters:

cur_iter (int) – Current iteration.
images (list[ArrayLikeFloat]) – Images to show.
image_names (list[str]) – Image names.
masks (list[ArrayLikeUInt]) – Segmentation masks to show, each with shape [H, W] or [N, H, W]. If the shape is [H, W], the mask is assumed to be a semantic segmentation mask with each pixel being the class id. If the shape is [N, H, W], each mask is assumed to be a binary mask with each pixel being either 0 or 1.
class_ids (list[ArrayLikeInt], optional) – Class ids for each mask, with shape [N]. If set, the masks are assumed to be binary masks and the length of class_ids must match the amount of masks. Defaults to None.

Return type:

None

process_single_image(image, image_name, masks, class_ids=None)[source]

Processes a single image entry.

Parameters:

image (ArrayLikeFloat) – Images to show.
image_name (str) – Name of the image.
masks (ArrayLikeUInt) – Binary masks to show, each with shape [N, H, W] or [H, W].
class_ids (ArrayLikeInt, optional) – Class ids for each mask, with shape [N]. Defaults to None.

Return type:

None

show(cur_iter, blocking=True)[source]

Shows the processed images in a interactive window.

Parameters:

cur_iter (int) – Current iteration.
blocking (bool) – If the visualizer should be blocking i.e. wait for human input for each image

Return type:

None

save_to_disk(cur_iter, output_folder)[source]

Saves the visualization to disk.

Writes all processes samples to the output folder naming each image <sample.image_name>.<filetype>.

Parameters:

cur_iter (int) – Current iteration.
output_folder (str) – Folder where the output should be written.

Return type:

None

Modules

`bbox3d_visualizer`	Bounding box 3D visualizer.
`bev_visualizer`	BEV Bounding box 3D visualizer.
`bounding_box_visualizer`	Bounding box visualizer.
`canvas`	Vis4D image canvas backends.
`functional`	Function interface for image visualization functions.
`seg_mask_visualizer`	Segmentation mask visualizer.
`util`	Utility functions for image processing operations.
`viewer`	Viewer implementations to display images.