| import sys |
| sys.path.append('droid_slam') |
|
|
| from tqdm import tqdm |
| import numpy as np |
| import torch |
| import lietorch |
| import cv2 |
| import os |
| import glob |
| import time |
| import argparse |
|
|
| import torch.nn.functional as F |
| from droid import Droid |
|
|
|
|
| def show_image(image): |
| image = image.permute(1, 2, 0).cpu().numpy() |
| cv2.imshow('image', image / 255.0) |
| cv2.waitKey(1) |
|
|
| def image_stream(datapath, image_size=[320, 512]): |
| """ image generator """ |
|
|
| fx, fy, cx, cy = 517.3, 516.5, 318.6, 255.3 |
|
|
| K_l = np.array([fx, 0.0, cx, 0.0, fy, cy, 0.0, 0.0, 1.0]).reshape(3,3) |
| d_l = np.array([0.2624, -0.9531, -0.0054, 0.0026, 1.1633]) |
|
|
| |
| images_list = sorted(glob.glob(os.path.join(datapath, 'rgb', '*.png')))[::2] |
| |
| for t, imfile in enumerate(images_list): |
| image = cv2.imread(imfile) |
| ht0, wd0, _ = image.shape |
| image = cv2.undistort(image, K_l, d_l) |
| image = cv2.resize(image, (320+32, 240+16)) |
| image = torch.from_numpy(image).permute(2,0,1) |
|
|
| intrinsics = torch.as_tensor([fx, fy, cx, cy]).cuda() |
| intrinsics[0] *= image.shape[2] / 640.0 |
| intrinsics[1] *= image.shape[1] / 480.0 |
| intrinsics[2] *= image.shape[2] / 640.0 |
| intrinsics[3] *= image.shape[1] / 480.0 |
|
|
| |
| intrinsics[2] -= 16 |
| intrinsics[3] -= 8 |
| image = image[:, 8:-8, 16:-16] |
|
|
| yield t, image[None], intrinsics |
|
|
| if __name__ == '__main__': |
| parser = argparse.ArgumentParser() |
| parser.add_argument("--datapath") |
| parser.add_argument("--weights", default="droid.pth") |
| parser.add_argument("--buffer", type=int, default=512) |
| parser.add_argument("--image_size", default=[240, 320]) |
| parser.add_argument("--disable_vis", action="store_true") |
|
|
| parser.add_argument("--beta", type=float, default=0.6) |
| parser.add_argument("--filter_thresh", type=float, default=1.75) |
| parser.add_argument("--warmup", type=int, default=12) |
| parser.add_argument("--keyframe_thresh", type=float, default=2.25) |
| parser.add_argument("--frontend_thresh", type=float, default=12.0) |
| parser.add_argument("--frontend_window", type=int, default=25) |
| parser.add_argument("--frontend_radius", type=int, default=2) |
| parser.add_argument("--frontend_nms", type=int, default=1) |
|
|
| parser.add_argument("--backend_thresh", type=float, default=15.0) |
| parser.add_argument("--backend_radius", type=int, default=2) |
| parser.add_argument("--backend_nms", type=int, default=3) |
| args = parser.parse_args() |
|
|
| args.stereo = False |
| torch.multiprocessing.set_start_method('spawn') |
|
|
| print("Running evaluation on {}".format(args.datapath)) |
| print(args) |
|
|
| droid = Droid(args) |
| time.sleep(5) |
|
|
| tstamps = [] |
| for (t, image, intrinsics) in tqdm(image_stream(args.datapath)): |
| if not args.disable_vis: |
| show_image(image) |
| droid.track(t, image, intrinsics=intrinsics) |
|
|
|
|
| traj_est = droid.terminate(image_stream(args.datapath)) |
|
|
| |
|
|
| print("#"*20 + " Results...") |
|
|
| import evo |
| from evo.core.trajectory import PoseTrajectory3D |
| from evo.tools import file_interface |
| from evo.core import sync |
| import evo.main_ape as main_ape |
| from evo.core.metrics import PoseRelation |
|
|
| image_path = os.path.join(args.datapath, 'rgb') |
| images_list = sorted(glob.glob(os.path.join(image_path, '*.png')))[::2] |
| tstamps = [float(x.split('/')[-1][:-4]) for x in images_list] |
|
|
| traj_est = PoseTrajectory3D( |
| positions_xyz=traj_est[:,:3], |
| orientations_quat_wxyz=traj_est[:,3:], |
| timestamps=np.array(tstamps)) |
|
|
| gt_file = os.path.join(args.datapath, 'groundtruth.txt') |
| traj_ref = file_interface.read_tum_trajectory_file(gt_file) |
|
|
| traj_ref, traj_est = sync.associate_trajectories(traj_ref, traj_est) |
| result = main_ape.ape(traj_ref, traj_est, est_name='traj', |
| pose_relation=PoseRelation.translation_part, align=True, correct_scale=True) |
|
|
|
|
| print(result) |
|
|
|
|
