def main():
parser = ArgumentParser()
parser.add_argument('pcd', help='Point cloud file')
parser.add_argument('config', help='Config file')
parser.add_argument('checkpoint', help='Checkpoint file')
parser.add_argument('--device',
default='cuda:0',
help='Device used for inference')
parser.add_argument('--score-thr',
type=float,
default=0.0,
help='bbox score threshold')
parser.add_argument('--out-dir',
type=str,
default='demo',
help='dir to save results')
parser.add_argument('--show',
action='store_true',
help='show online visuliaztion results')
parser.add_argument('--snapshot',
action='store_true',
help='whether to save online visuliaztion results')
args = parser.parse_args()
# build the model from a config file and a checkpoint file
model = init_detector(args.config, args.checkpoint, device=args.device)
# test a single image
result, data = inference_detector(model, args.pcd)
# show the results
show_result_meshlab(data,
result,
args.out_dir,
args.score_thr,
show=args.show,
snapshot=args.snapshot)
def test_inference_detector():
pcd = 'tests/data/kitti/training/velodyne_reduced/000000.bin'
detector_cfg = 'configs/pointpillars/hv_pointpillars_secfpn_' \
'6x8_160e_kitti-3d-3class.py'
detector = init_detector(detector_cfg, device='cpu')
results = inference_detector(detector, pcd)
bboxes_3d = results[0][0]['boxes_3d']
scores_3d = results[0][0]['scores_3d']
labels_3d = results[0][0]['labels_3d']
assert bboxes_3d.tensor.shape[0] >= 0
assert bboxes_3d.tensor.shape[1] == 7
assert scores_3d.shape[0] >= 0
assert labels_3d.shape[0] >= 0
def test_inference_detector():
if not torch.cuda.is_available():
pytest.skip('test requires GPU and torch+cuda')
pcd = 'tests/data/kitti/training/velodyne_reduced/000000.bin'
detector_cfg = 'configs/pointpillars/hv_pointpillars_secfpn_' \
'6x8_160e_kitti-3d-3class.py'
detector = init_model(detector_cfg, device='cuda:0')
results = inference_detector(detector, pcd)
bboxes_3d = results[0][0]['boxes_3d']
scores_3d = results[0][0]['scores_3d']
labels_3d = results[0][0]['labels_3d']
assert bboxes_3d.tensor.shape[0] >= 0
assert bboxes_3d.tensor.shape[1] == 7
assert scores_3d.shape[0] >= 0
assert labels_3d.shape[0] >= 0
def run(self, data_bin, threshold=0.3):
result, data = inference_detector(self.model, data_bin)
obj_ind = result[0]['scores_3d'] >= threshold
pred_3d = result[0]['boxes_3d'].corners[obj_ind, ...]
pred_2d = []
for obj in pred_3d:
obj_2d = self.calib.project_velo_to_image(obj)
pred_2d.append([
np.min(obj_2d, axis=0)[0],
np.min(obj_2d, axis=0)[1],
np.max(obj_2d, axis=0)[0],
np.max(obj_2d, axis=0)[1]
])
return pred_2d, [
xx.squeeze() for xx in np.split(pred_3d, pred_3d.shape[0], axis=0)
]
def main():
parser = ArgumentParser()
parser.add_argument('pcd', help='Point cloud file')
parser.add_argument('config', help='Config file')
parser.add_argument('checkpoint', help='Checkpoint file')
parser.add_argument(
'--device', default='cuda:0', help='Device used for inference')
parser.add_argument(
'--score-thr', type=float, default=0.0, help='bbox score threshold')
parser.add_argument(
'--out-dir', type=str, default='demo', help='dir to save results')
args = parser.parse_args()
# build the model from a config file and a checkpoint file
model = init_detector(args.config, args.checkpoint, device=args.device)
# test a single image
result, data = inference_detector(model, args.pcd)
def execution(idx):
global dataset, model, cfg, args
print()
# Select and infer a sample
sample = dataset.__getitem__(idx)
if isinstance(sample['img_metas'], list):
filename = sample['img_metas'][0].data['pts_filename']
else:
filename = sample['img_metas'].data['pts_filename']
pcd_file = os.path.join(dataset.root_split, dataset.pts_prefix, filename)
print("Selected pcd_file:", pcd_file)
result, data = inference_detector(model, pcd_file)
print("Number of detected bboxes:", len(result['labels_3d']))
# Convert to MeshLab
os.makedirs(args.outdir, exist_ok=True)
show_result_meshlab(data, result, args.outdir)
print("Visualization result is saved in \'{}\'. "
"Please use MeshLab to visualize it.".format(
os.path.join(args.outdir,
os.path.basename(pcd_file).split('.')[0])))
from mmdet3d.apis import init_detector, inference_detector, show_result_meshlab #%% config_file = '../configs/3dssd/3dssd_kitti-3d-car.py' # download the checkpoint from model zoo and put it in `checkpoints/` checkpoint_file = '../checkpoints/3dssd_kitti-3d-car/latest.pth' #%% # build the model from a config file and a checkpoint file model = init_detector(config_file, checkpoint_file, device='cuda:0') #%% # test a single sample pcd = 'kitti_000008.bin' result, data = inference_detector(model, pcd) print(model)
def main():
parser = ArgumentParser()
parser.add_argument(
'--pcd',
type=str,
default='./demo/kitti_000008.bin',
help='Point cloud file'
) #/data/cmpe249-fa21/4c_train5678/training/velodyne/008118.bin
parser.add_argument(
'--config',
type=str,
default=
'./configs/second/myhv_second_secfpn_sbn_2x16_2x_waymoD5-3d-4class.py',
help='Config file')
parser.add_argument(
'--checkpoint',
type=str,
default=
'../modelzoo_mmdetection3d/hv_second_secfpn_6x8_80e_kitti-3d-3class_20200620_230238-9208083a.pth',
help='Checkpoint file')
parser.add_argument('--device',
default='cuda:0',
help='Device used for inference')
parser.add_argument('--score-thr',
type=float,
default=0.6,
help='bbox score threshold')
parser.add_argument('--out-dir',
type=str,
default='demo',
help='dir to save results')
args = parser.parse_args()
# build the model from a config file and a checkpoint file
model = init_detector(args.config, args.checkpoint, device=args.device)
# test a single image
result, data = inference_detector(model, args.pcd)
print(result) # three keys: 'boxes_3d', 'scores_3d', 'labels_3d'
points = data['points'][0][0].cpu().numpy() # points number *4
pts_filename = data['img_metas'][0][0]['pts_filename']
print("pts_filename:", pts_filename)
file_name = osp.split(pts_filename)[-1].split('.')[0] #006767
print(data['img_metas'])
print(data['img_metas'][0][0]['box_mode_3d']) #Box3DMode.LIDAR
print("results len:", len(result[0])) # len=1,
for res in result[0].keys():
print(res)
#boxes_3d, scores_3d, labels_3d
pred_bboxes = result[0]['boxes_3d'].tensor.numpy()
print(
pred_bboxes
) # [11,7], Each row is (x, y, z, x_size, y_size, z_size, yaw) in Box3DMode.LIDAR
print(type(pred_bboxes)) #numpy.ndarray
print("box_mode_3d:",
data['img_metas'][0][0]['box_mode_3d']) #Box3DMode.LIDAR
# show the results (points save to xxx_points.obj file, pred3d box save to xxx_pred.ply, these two files can opend by meshlab)
show_result_meshlab(data, result, args.out_dir)
img_pts = np.matmul(corners3d_hom, P2.T) # (N, 8, 3), The transposed array.
x, y = img_pts[:, :, 0] / img_pts[:, :, 2], img_pts[:, :, 1] / img_pts[:, :, 2]
x1, y1 = np.min(x, axis=1), np.min(y, axis=1)
x2, y2 = np.max(x, axis=1), np.max(y, axis=1)
boxes = np.concatenate((x1.reshape(-1, 1), y1.reshape(-1, 1), x2.reshape(-1, 1), y2.reshape(-1, 1)), axis=1)
boxes_corner = np.concatenate((x.reshape(-1, 8, 1), y.reshape(-1, 8, 1)), axis=2)
return boxes, boxes_corner
# build the model from a config file and a checkpoint file
model = init_detector(base_folder+config_file, base_folder+checkpoint_file, device='cuda:0')
# test a single image and show the results
result, data = inference_detector(model, point_cloud)
points = data['points'][0][0].cpu().numpy()# points number *4
pts_filename = data['img_metas'][0][0]['pts_filename']
file_name = osp.split(pts_filename)[-1].split('.')[0] #006767
print(data['img_metas'])
print(data['img_metas'][0][0]['box_mode_3d']) #Box3DMode.LIDAR
print(len(result))
for res in result[0].keys():
print(res)
pred_bboxes = result[0]['boxes_3d'].tensor.numpy()
print(pred_bboxes)# [11,7], Each row is (x, y, z, x_size, y_size, z_size, yaw) in Box3DMode.LIDAR
print(type(pred_bboxes))#numpy.ndarray
# boxes_3d
# scores_3d
if __name__ == '__main__':
mmdet3d_dir = '../mmdetection3d/'
config_file = os.path.join(mmdet3d_dir,
"configs/3dssd/3dssd_kitti-3d-car.py")
ckpt_file = os.path.join(mmdet3d_dir,
"checkpoints/3dssd_kitti-3d-car/latest.pth")
# build the model from a config file and a checkpoint file
from mmdet3d.apis import init_detector, inference_detector, show_result_meshlab
from tqdm import tqdm
model = init_detector(config_file, ckpt_file, device='cuda:0')
preds = []
for i in tqdm(range(len(dataset))):
result, data = inference_detector(model, dataset.get_lidar_filename(i))
threshold = 0.5
bboxes, scores = result[0]['boxes_3d'].corners, result[0]['scores_3d']
pred = bboxes[scores > threshold]
preds.append(pred)
def anim(t):
global fig_3d
mlab.clf()
t = int(t)
img = dataset.get_image(t)
pcd = dataset.get_lidar(t)
calib = dataset.get_calibration(t)
_, gt, _, _ = dataset.get_object(t)
pred = preds[t]
img_height, img_width, _ = img.shape
