| 123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657 |
- from scene.cameras import Camera
- import numpy as np
- from utils.general_utils import PILtoTorch
- from utils.graphics_utils import fov2focal
- def loadCam(args, id, cam_info, resolution_scale):
- orig_w, orig_h = cam_info.image.size
- if args.resolution in [1, 2, 4, 8]:
- resolution = round(orig_w/(resolution_scale * args.resolution)), round(orig_h/(resolution_scale * args.resolution))
- else: # should be a type that converts to float
- global_down = orig_w/args.resolution
- scale = float(global_down) * float(resolution_scale)
- resolution = (int(orig_w / scale), int(orig_h / scale))
- resized_image_rgb = PILtoTorch(cam_info.image, resolution)
- gt_image = resized_image_rgb[:3, ...]
- loaded_mask = None
- if resized_image_rgb.shape[1] == 4:
- loaded_mask = resized_image_rgb[3:4, ...]
- return Camera(colmap_id=cam_info.uid, R=cam_info.R, T=cam_info.T,
- FoVx=cam_info.FovX, FoVy=cam_info.FovY,
- image=gt_image, gt_alpha_mask=loaded_mask,
- image_name=cam_info.image_name, uid=id)
- def cameraList_from_camInfos(cam_infos, resolution_scale, args):
- camera_list = []
- for id, c in enumerate(cam_infos):
- camera_list.append(loadCam(args, id, c, resolution_scale))
- return camera_list
- def camera_to_JSON(id, camera : Camera):
- Rt = np.zeros((4, 4))
- Rt[:3, :3] = camera.R.transpose()
- Rt[:3, 3] = camera.T
- Rt[3, 3] = 1.0
- W2C = np.linalg.inv(Rt)
- pos = W2C[:3, 3]
- rot = W2C[:3, :3]
- serializable_array_2d = [x.tolist() for x in rot]
- camera_entry = {
- 'id' : id,
- 'img_name' : camera.image_name,
- 'width' : camera.width,
- 'height' : camera.height,
- 'position': pos.tolist(),
- 'rotation': serializable_array_2d,
- 'fy' : fov2focal(camera.FovY, camera.height),
- 'fx' : fov2focal(camera.FovX, camera.width)
- }
- return camera_entry
|
