How do I read a pcd (point cloud data) file using python and convert point cloud to depth image?

1

I now have a pcd profile, I hope I can read it using python.
But I have not learned anything about pcl. I don’t know how to convert this data into a depth image.
Hope someone can help me with it.
I am using version python 3.6. The data set is Cornell’s grasp dataset.

2

Usually you would have to project the point cloud onto an image plane using the focal lengths and center position. Here is some code for the PCL.

3

I have implemented it on C++.
But I don’t know how to implement it in Python. Can I use PCL in a python environment?

4

Apparently there are some PCL Python packages:

5

I implemented it on python.

import os
import math
import numpy as np
from PIL import Image


pcd_path = "input_path"
pcd_name = "pcd"
pcd_mat = ".txt"
pcd_number = 1000


with open(pcd_path+ pcd_name+str(pcd_number) + pcd_mat, "r") as pcd_file:
    lines = [line.strip().split(" ") for line in pcd_file.readlines()]

img_height = 480
img_width = 640
is_data = False
min_d = 0
max_d = 0
img_depth = np.zeros((img_height, img_width), dtype='f8')
for line in lines:
    if line[0] == 'DATA':  # skip the header
        is_data = True
        continue
    if is_data:
        d = max(0., float(line[2]))
        i = int(line[4])
        col = i % img_width
        row = math.floor(i / img_width)
        img_depth[row, col] = d
        min_d = min(d, min_d)
        max_d = max(d, max_d)

max_min_diff = max_d - min_d


def normalize(x):
    return 255 * (x - min_d) / max_min_diff
normalize = np.vectorize(normalize, otypes=[np.float])
img_depth = normalize(img_depth)
img_depth_file = Image.fromarray(img_depth)
img_depth_file.convert('RGB').save(os.path.join(output_path,pcd_name+str(pcd_number)+'_depth_image.png'))

refer : https://stackoverflow.com/questions/42430479/how-to-extract-depth-information-from-the-3d-point-cloud-data?fbclid=IwAR37ukkHd2_dp0tV8g-oRiXGHguSHbXG8YbmMwn8ZPeACxkAU2jXbPnmpc8

6

I found this python file and test ok, you can use this file to load pcd file just like pcl

1 Like