search

liulei01 / DRBox

A deep learning based algorithm to detect rotated object, for example, objects in remote sensing images
Other
421 stars 142 forks source link

CUDNN and OpenCV version #6

Open arasharchor opened 6 years ago

arasharchor commented 6 years ago

Hi,

If I want to use images with different resolution e.g. 15 cm. what should I change?

I have changed "resolution" parameter in > SelectRotatedTarget.m to 0.15. Also in deploy.py file I have changed "resolutionIn = 0.15 and resolutionOut = [0.15]. The results are however so bad only a few cars are detected with mostly wrong rotation estimation. I am using a 569x561px image.

BTW, I am trying to install your modified caffe with CUDNN and OpenCV. Could you please let me know which version you used?

keck91 commented 5 years ago

BTW, I am trying to install your modified caffe with CUDNN and OpenCV. Could you please let me know which version you used?

CUDA 8.0, openCV 2.4.9 & cuDNN 5.1 worked for me

Viktor-Paul commented 3 years ago

Hi,

If I want to use images with different resolution e.g. 15 cm. what should I change?

I have changed "resolution" parameter in > SelectRotatedTarget.m to 0.15. Also in deploy.py file I have changed "resolutionIn = 0.15 and resolutionOut = [0.15]. The results are however so bad only a few cars are detected with mostly wrong rotation estimation. I am using a 569x561px image.

BTW, I am trying to install your modified caffe with CUDNN and OpenCV. Could you please let me know which version you used?

About detecting the vehicle, is the dataset you used your own dataset?