相机 (Camera)

class genesis.vis.camera.Camera(visualizer, idx=0, model='pinhole', res=(320, 320), pos=(0.5, 2.5, 3.5), lookat=(0.5, 0.5, 0.5), up=(0.0, 0.0, 1.0), fov=30, aperture=2.8, focus_dist=None, GUI=False, spp=256, denoise=True, near=0.05, far=100.0, transform=None)

基类：RBC

Genesis camera class. The camera can be used to render RGB, depth, and segmentation images. The camera can use either rasterizer or raytracer for rendering, specified by scene.renderer. The camera also comes with handy tools such as video recording.

参数:

• visualizer (genesis.Visualizer) – The visualizer object that the camera is associated with.

• idx (int) – The index of the camera.

• model (str) – Specifies the camera model. Options are ‘pinhole’ or ‘thinlens’.

• res (tuple of int, shape (2,)) – The resolution of the camera, specified as a tuple (width, height).

• pos (tuple of float, shape (3,)) – The position of the camera in the scene, specified as (x, y, z).

• lookat (tuple of float, shape (3,)) – The point in the scene that the camera is looking at, specified as (x, y, z).

• up (tuple of float, shape (3,)) – The up vector of the camera, defining its orientation, specified as (x, y, z).

• fov (float) – The vertical field of view of the camera in degrees.

• aperture (float) – The aperture size of the camera, controlling depth of field.

• focus_dist (float | None) – The focus distance of the camera. If None, it will be auto-computed using pos and lookat.

• GUI (bool) – Whether to display the camera’s rendered image in a separate GUI window.

• spp (int, optional) – Samples per pixel. Defaults to 256.

• denoise (bool) – Whether to denoise the camera’s rendered image.

• near (float) – The near plane of the camera.

• far (float) – The far plane of the camera.

• transform (np.ndarray, shape (4, 4), optional) – The transform matrix of the camera.

render(rgb=True, depth=False, segmentation=False, colorize_seg=False, normal=False)

Render the camera view. Note that the segmentation mask can be colorized, and if not colorized, it will store an object index in each pixel based on the segmentation level specified in vis_options.segmentation_level. For example, if segmentation_level=’link’, the segmentation mask will store link_idx, which can then be used to retrieve the actual link objects using scene.rigid_solver.links[link_idx].

参数:

• rgb (bool, optional) – Whether to render an RGB image.

• depth (bool, optional) – Whether to render a depth image.

• segmentation (bool, optional) – Whether to render the segmentation mask.

• colorize_seg (bool, optional) – If True, the segmentation mask will be colorized.

• normal (bool, optional) – Whether to render the surface normal.

返回:

• rgb_arr (np.ndarray) – The rendered RGB image.

• depth_arr (np.ndarray) – The rendered depth image.

• seg_arr (np.ndarray) – The rendered segmentation mask.

• normal_arr (np.ndarray) – The rendered surface normal.

set_pose(transform=None, pos=None, lookat=None, up=None)

Set the pose of the camera. Note that transform has a higher priority than pos, lookat, and up. If transform is provided, the camera pose will be set based on the transform matrix. Otherwise, the camera pose will be set based on pos, lookat, and up.

参数:

• transform (np.ndarray, shape (4, 4), optional) – The transform matrix of the camera.

• pos (array-like, shape (3,), optional) – The position of the camera.

• lookat (array-like, shape (3,), optional) – The lookat point of the camera.

• up (array-like, shape (3,), optional) – The up vector of the camera.

set_params(fov=None, aperture=None, focus_dist=None)

Update the camera parameters.

参数:

• fov (float, optional) – The vertical field of view of the camera.

• aperture (float, optional) – The aperture of the camera. Only supports ‘thinlens’ camera model.

• focus_dist (float, optional) – The focus distance of the camera. Only supports ‘thinlens’ camera model.

start_recording()

Start recording on the camera. After recording is started, all the rgb images rendered by camera.render() will be stored, and saved to a video file when camera.stop_recording() is called.

pause_recording()

Pause recording on the camera. After recording is paused, the rgb images rendered by camera.render() will not be stored. Recording can be resumed by calling camera.start_recording() again.

stop_recording(save_to_filename=None, fps=60)

Stop recording on the camera. Once this is called, all the rgb images stored so far will be saved to a video file. If save_to_filename is None, the video file will be saved with the name ‘{caller_file_name}_cam_{camera.idx}.mp4’.

参数:

• save_to_filename (str, optional) – Name of the output video file. If not provided, the name will be default to the name of the caller file, with camera idx, a timestamp and ‘.mp4’ extension.

• fps (int, optional) – The frames per second of the video file.

property is_built

Whether the camera is built.

property idx

The integer index of the camera.

property uid

The unique ID of the camera

property model

pinhole or thinlens.

Type:

The camera model

property res

The resolution of the camera.

property fov

The field of view of the camera.

property aperture

The aperture of the camera.

property focal_len

The focal length for thinlens camera. Returns -1 for pinhole camera.

property focus_dist

The focus distance of the camera.

property GUI

Whether the camera will display the rendered images in a separate window.

property spp

Samples per pixel of the camera.

property denoise

Whether the camera will denoise the rendered image in raytracer.

property near

The near plane of the camera.

property far

The far plane of the camera.

property aspect_ratio

The aspect ratio of the camera.

property pos

The current position of the camera.

property lookat

The current lookat point of the camera.

property up

The current up vector of the camera.

property transform

The current transform matrix of the camera.

property extrinsics

The current extrinsics matrix of the camera.

property intrinsics

The current intrinsics matrix of the camera.