jiang-muyun / pointnet12 Goto Github PK
View Code? Open in Web Editor NEWCombination of Pointnet and Pointnet++ with KITTI Point Cloud Semantic Segmentation Dataset
Combination of Pointnet and Pointnet++ with KITTI Point Cloud Semantic Segmentation Dataset
@Jiang-Muyun
the code in data_utils/kitti_utils.py erase the 4th column
def calib_cam2cam(self, fn_c2c, mode = '02'):
"""
If your image is 'rectified image' :get only Projection(P : 3x4) matrix is enough
but if your image is 'distorted image'(not rectified image) :
you need undistortion step using distortion coefficients(5 : D)
In this code, only P matrix info is used for rectified image
"""
# with open(fn_c2c, "r") as f: c2c_file = f.readlines()
for line in open(fn_c2c, "r"):
(key, val) = line.split(':', 1)
if key == ('P_rect_' + mode):
P = np.fromstring(val, sep=' ')
P = P.reshape(3, 4)
P = P[:3, :3] # erase 4th column ([0,0,0])
return P
but the 4th column of P_rect_02 in config/calib_cam_to_cam.txt is not 0.
P_rect_02:
7.215377e+02 0.000000e+00 6.095593e+02 4.485728e+01
0.000000e+00 7.215377e+02 1.728540e+02 2.163791e-01
0.000000e+00 0.000000e+00 1.000000e+00 2.745884e-03
the wrong information is like below,thks!
Traceback (most recent call last):
File "pcdvis.py", line 154, in <module>
vis(args)
File "pcdvis.py", line 116, in vis
point_cloud, label = kitti_utils.get(part, index, load_image=True)
File "/home/gz/桌面/3DCVcode/segmentation/PointNet12KiTTI/data_utils/kitti_utils.py", line 190, in get
assert os.path.exists(fn_frame), 'Broken dataset %s' % (fn_frame)
AssertionError: Broken dataset /home/gz/桌面/3DCVcode/segmentation/SemanticKITTI/data/dataset/dataset/sequences/01/image_2/000000.png
the wrong information is like below,thks!
Traceback (most recent call last):
File "pcdvis.py", line 154, in
vis(args)
File "pcdvis.py", line 107, in vis
vis_handle = Window_Manager()
File "pcdvis.py", line 35, in init
self.vis.create_window(width=800, height=800, left=100)
RuntimeError: [Open3D ERROR] GLFW Error: GLX: Failed to create context: BadLength (poly request too large or internal Xlib length error)
I set the width and height to 400, but it doesn't work .
Your project is great. As a novice, I want to know which file semseg.py and file pcdseg.py is running the semantic segmentation program.
Hi,
Could you please upload the checkpoints folder with the pre-trained model files. These files are mentioned in the script pcdvis.py.
Broken dataset D:/KT/odometry/dataset/sequences/01/velodyne/000000.bin
I checked my dataset and there is no velodyne files.
The dataset i'm using is [Download odometry data set (color, 65 GB)],Download odometry data set (velodyne laser data, 80 GB)
and symantic KITTI dataset.
After combining these datasets, it's finally running.
Hello, I am a new student, I saw the configuration instructions you left and would like to know in more detail about your version of pytorch and other configurations.
Hi Muyun, thank you so much for your work. I am running the code but I found that the test result for PointNet is too high and its miou reaches around 0.47. I know this is a project you did quite a long time ago, but I am wondering if you could remember something regarding the result you got? Any updates would be appreciated. Thanks!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.