hofee/nbv_reconstruction
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update preprocessor

Browse Source
This commit is contained in:
hofee 2024-10-03 23:36:18 +08:00
parent d7561738c6
commit fd7614c847
4 changed files with 58 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
configs/local/view_generate_config.yaml
View File

@ -7,16 +7,16 @@ runner:
name: debug
root_dir: experiments
generate:
port: 5003
from: 3000
to: -1 # -1 means all
port: 5002
from: 2200
to: 2300 # -1 means all
object_dir: /media/hofee/data/data/scaled_object_meshes
table_model_path: /media/hofee/data/data/others/table.obj
output_dir: /media/hofee/repository/new_full_data
output_dir: /media/hofee/repository/new_data_with_normal
binocular_vision: true
plane_size: 10
max_views: 512
min_views: 64
min_views: 128
random_view_ratio: 0.2
min_cam_table_included_degree: 20
max_diag: 0.7

58
preprocess/preprocessor.py
View File

@ -1,6 +1,13 @@
import os
import json
import numpy as np
import sys
np.random.seed(0)
# append parent directory to sys.path
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
print(sys.path)
from utils.reconstruction import ReconstructionUtil
from utils.data_load import DataLoadUtil
from utils.pts import PtsUtil
@ -55,8 +62,7 @@ def get_world_points(depth, cam_intrinsic, cam_extrinsic):
points_camera_world = np.dot(cam_extrinsic, points_camera_aug.T).T[:, :3]
return points_camera_world
def get_world_normals(normal_image, cam_extrinsic):
normals = normal_image.reshape(-1, 3)
def get_world_normals(normals, cam_extrinsic):
normals = normals / np.linalg.norm(normals, axis=1, keepdims=True)
normals_world = np.dot(cam_extrinsic[:3, :3], normals.T).T
return normals_world
@ -75,7 +81,7 @@ def get_scan_points_indices(scan_points, mask, display_table_mask_label, cam_int
return selected_points_indices
def save_scene_data(root, scene):
def save_scene_data(root, scene, scene_idx=0, scene_total=1):
''' configuration '''
target_mask_label = (0, 255, 0, 255)
@ -88,11 +94,12 @@ def save_scene_data(root, scene):
''' scan points '''
display_table_info = DataLoadUtil.get_display_table_info(root, scene)
radius = display_table_info["radius"]
scan_points = ReconstructionUtil.generate_scan_points(display_table_top=0,display_table_radius=radius)
scan_points = np.asarray(ReconstructionUtil.generate_scan_points(display_table_top=0,display_table_radius=radius))
''' read frame data(depth|mask|normal) '''
frame_num = DataLoadUtil.get_scene_seq_length(root, scene)
for frame_id in range(frame_num):
print(f"[scene({scene_idx}/{scene_total})|frame({frame_id}/{frame_num})]Processing {scene} frame {frame_id}")
path = DataLoadUtil.get_path(root, scene, frame_id)
cam_info = DataLoadUtil.load_cam_info(path, binocular=True)
depth_L, depth_R = DataLoadUtil.load_depth(
@ -104,30 +111,36 @@ def save_scene_data(root, scene):
normal_L = DataLoadUtil.load_normal(path, binocular=True, left_only=True)
''' scene points '''
scene_points_L = get_world_points(depth_L, cam_info["cam_intrinsic_L"], cam_info["cam_extrinsic_L"])
scene_points_R = get_world_points(depth_R, cam_info["cam_intrinsic_R"], cam_info["cam_extrinsic_R"])
scene_points_L, random_sample_idx_L = PtsUtil.random_downsample_point_cloud(
scene_points_L = get_world_points(depth_L, cam_info["cam_intrinsic"], cam_info["cam_to_world"])
scene_points_R = get_world_points(depth_R, cam_info["cam_intrinsic"], cam_info["cam_to_world_R"])
sampled_scene_points_L, random_sample_idx_L = PtsUtil.random_downsample_point_cloud(
scene_points_L, random_downsample_N, require_idx=True
)
scene_points_R = PtsUtil.random_downsample_point_cloud(
sampled_scene_points_R = PtsUtil.random_downsample_point_cloud(
scene_points_R, random_downsample_N
)
scene_overlap_points, overlap_idx_L = PtsUtil.get_overlapping_points(
scene_points_L, scene_points_R, voxel_size, require_idx=True
sampled_scene_points_L, sampled_scene_points_R, voxel_size, require_idx=True
)
if scene_overlap_points.shape[0] < 1024:
scene_overlap_points = sampled_scene_points_L
overlap_idx_L = np.arange(sampled_scene_points_L.shape[0])
train_input_points, train_input_idx = PtsUtil.random_downsample_point_cloud(
scene_overlap_points, train_input_pts_num, require_idx=True
)
''' target points '''
mask_img = mask_L
mask_L = mask_L.reshape(-1, 4)
mask_L = (mask_L == target_mask_label).all(axis=-1)
mask_overlap = mask_L[random_sample_idx_L][overlap_idx_L]
scene_normals_L = get_world_normals(normal_L, cam_info["cam_extrinsic_L"])
scene_normals_L = normal_L.reshape(-1, 3)
target_overlap_normals = scene_normals_L[random_sample_idx_L][overlap_idx_L][mask_overlap]
target_normals = get_world_normals(target_overlap_normals, cam_info["cam_to_world"])
target_points = scene_overlap_points[mask_overlap]
target_normals = scene_normals_L[mask_overlap]
filtered_target_points, filtered_idx = PtsUtil.filter_points(
target_points, target_normals, cam_info["cam_extrinsic_L"], filter_degree, require_idx=True
target_points, target_normals, cam_info["cam_to_world"], filter_degree, require_idx=True
)
''' train_input_mask '''
@ -135,7 +148,7 @@ def save_scene_data(root, scene):
''' scan points indices '''
scan_points_indices = get_scan_points_indices(scan_points, mask_L, display_table_mask_label, cam_info["cam_intrinsic_L"], cam_info["cam_extrinsic_L"])
scan_points_indices = get_scan_points_indices(scan_points, mask_img, display_table_mask_label, cam_info["cam_intrinsic"], cam_info["cam_to_world"])
save_full_points(root, scene, frame_id, train_input_points)
save_target_points(root, scene, frame_id, filtered_target_points)
@ -146,6 +159,19 @@ def save_scene_data(root, scene):
if __name__ == "__main__":
root = ""
for scene in os.listdir(root):
save_scene_data(root, scene)
#root = "/media/hofee/repository/new_data_with_normal"
root = "/media/hofee/repository/test_sample"
list_path = "/media/hofee/repository/test_sample/test_sample_list.txt"
scene_list = []
with open(list_path, "r") as f:
for line in f:
scene_list.append(line.strip())
from_idx = 0
to_idx = len(scene_list)
cnt = 0
total = to_idx - from_idx
for scene in scene_list[from_idx:to_idx]:
save_scene_data(root, scene, cnt, total)
cnt+=1

11
utils/data_load.py
View File

@ -190,7 +190,7 @@ class DataLoadUtil:
mask_path = os.path.join(
os.path.dirname(path), "mask", os.path.basename(path) + ".png"
)
mask_image = cv2.imread(mask_path, cv2.IMREAD_GRAYSCALE)
mask_image = cv2.imread(mask_path, cv2.IMREAD_UNCHANGED)
return mask_image
@staticmethod
@ -199,23 +199,22 @@ class DataLoadUtil:
normal_path_L = os.path.join(
os.path.dirname(path), "normal", os.path.basename(path) + "_L.png"
)
normal_image_L = cv2.imread(normal_path_L, cv2.IMREAD_UNCHANGED)
normal_image_L = cv2.imread(normal_path_L, cv2.IMREAD_COLOR)
normal_path_R = os.path.join(
os.path.dirname(path), "normal", os.path.basename(path) + "_R.png"
)
normal_image_R = cv2.imread(normal_path_R, cv2.IMREAD_UNCHANGED)
return normal_image_L, normal_image_R
normal_image_R = cv2.imread(normal_path_R, cv2.IMREAD_COLOR)
return normal_image_L[:3,:3], normal_image_R[:3,:3]
else:
if binocular and left_only:
normal_path = os.path.join(
os.path.dirname(path), "normal", os.path.basename(path) + "_L.png"
)
else:
normal_path = os.path.join(
os.path.dirname(path), "normal", os.path.basename(path) + ".png"
)
normal_image = cv2.imread(normal_path, cv2.IMREAD_UNCHANGED)
normal_image = cv2.imread(normal_path, cv2.IMREAD_COLOR)
return normal_image
@staticmethod

6
utils/pts.py
View File

@ -20,6 +20,8 @@ class PtsUtil:
@staticmethod
def random_downsample_point_cloud(point_cloud, num_points, require_idx=False):
if point_cloud.shape[0] == 0:
if require_idx:
return point_cloud, np.array([])
return point_cloud
idx = np.random.choice(len(point_cloud), num_points, replace=True)
if require_idx:
@ -60,7 +62,11 @@ class PtsUtil:
cos_theta = np.dot(normals_normalized, camera_axis)
theta_rad = np.deg2rad(theta)
idx = cos_theta > np.cos(theta_rad)
print(cos_theta, theta_rad)
filtered_points= points[idx]
# ------ Debug Start ------
import ipdb;ipdb.set_trace()
# ------ Debug End ------
if require_idx:
return filtered_points, idx
return filtered_points