使用训练好的.h5模型

https://blog.csdn.net/u010299280/article/details/82852727
AttributeError: ‘Model’ object has no attribute ‘predict_classes’
predict = model.predict(img)

安装CUDA 及CuDNN


安装cuda

https://zhuanlan.zhihu.com/p/94220564#:~:text=win10%E5%AE%89%E8%A3%85CUDA%E5%92%8CcuDNN%E7%9A%84%E6%AD%A3%E7%A1%AE%E5%A7%BF%E5%8A%BF%201%20%E3%80%81%E6%9F%A5%E7%9C%8B%E6%9C%AC%E6%9C%BA%E7%9A%84CUDA%E9%A9%B1%E5%8A%A8%E9%80%82%E9%85%8D%E7%89%88%E6%9C%AC%202%20%E3%80%81%E4%B8%8B%E8%BD%BDCUDA%E5%92%8CcuDNN%203,%E3%80%81%E5%AE%89%E8%A3%85CUDA%E5%92%8CcuDNN%204%20%E3%80%81%E6%B7%BB%E5%8A%A0%E7%8E%AF%E5%A2%83%E5%8F%98%E9%87%8F.C%3A%5CProgram%20Files%5CNVI%20…%205%20%E3%80%81%E6%A3%80%E6%9F%A5%E5%AE%89%E8%A3%85%E7%BB%93%E6%9E%9C

安装cuda与cudnn

https://blog.csdn.net/jhsignal/article/details/111401628

RCAN环境相关

CUDA 10.0 and cuDNN 7.4.2

Requireme

h5py<3.0
keras==2.2.4
numexpr
numpy>=1.18,!=1.19.4
tensorflow_gpu==1.13.1
tifffile
tqdm>=4.41.0

Anaconda

conda create -n tf python=3.7
pip install tensorflow-gpu==1.13.1

apply model

python apply.py -m **"E:**\\**code**\\**deepL**\\**3D-RCAN**\\**model**\\**" **-i **"E:**\\**download**\\**edge_download**\\**5%image000010s.tif" **-o**"E:**\\**download**\\**edge_download**\\**5%image000010out.tif"**

保存keras的模型

model.save("my_h5_model.h5")