[Open3D] Vertex normal estimation

1. 계산

print("Recompute the normal of the downsampled point cloud")

#finds adjacent points and calculate the principal axis of the adjacent points using covariance analysis.
estimate_normals(downpcd, 
    search_param = KDTreeSearchParamHybrid(radius = 0.1, max_nn = 30)) 
    # search radius and maximum nearest neighbor

draw_geometries([downpcd])

2. 접근

print("Print a normal vector of the 0th point")
print(downpcd.normals[0])

#>Print a normal vector of 0th point
#>[-0.27566603 -0.89197839 -0.35830543]

# np변환 

print("Print the normal vectors of the first 10 points")
np.asarray(downpcd.normals)[:10,:]

import numpy as np
from open3d import *

pcd = read_point_cloud("/workspace/tmp/normal/fragment.pcd")
# Instead of "setRadiusSearch()", you can make use of "setKSearch()", which takes an integer, K.
KD = KDTreeSearchParamHybrid(radius = 0.1, max_nn = 30)
estimate_normals(pcd, search_param = KD)
# setViewPoint() # 옵

print(pcd.has_normals())
pcd.normals[0]
pcd.points[0]

normals = np.asarray(pcd.normals)
normals[0]