To overcome the problem of the COVID-19 pandemic, we developed a detector based on the Convolutional Neural Network (CNN) using X-ray images. The architecture in the CNN method is composed of 32 filters, 2 convection layers and 1 dense layer using the Python Programming language. The results of this technology trial are able to recognize the symptoms of COVID-19 accurately with an error rate reaching 9.27 e-4% and an accuracy value reaching 100%. So with this technology can make it easier to identify the symptoms of COVID-19 in patients with suspected.
- Download the code
- Extract the code
- Edit your path of training data and save directory for the result categories of labels and features at preprocessing.py
- Edit your path for saving train data module at train_data.py
- Edit your path for testing data at test_data.py
- Open Command Prompt
- run python preprocessing.py
- run python train_data.py
- run python test_data.py
- python 3.6.8 version
- matplotlib 3.2.1 version
- opencv-python 4.2.0.32 version
- tensorflow 1.14.0 version
You can learn about python programming at https://pythonprogramming.net/