dogusyuksel / FireDetectionCNN

This project detects fire from images. It uses Convolutional Neural Network and predefined models to extract features from images.
MIT License
15 stars 6 forks source link

FireDetectionCNN

This project detects fire from images. It uses Convolutional Neural Network and predefined models to extract features from images.

To train modal;

  1. put your feeding images in the related part under the "dataset" folder
  2. open "train_model.py" and insert your parameters like model type, batch size etc (upside of the code)
  3. Run the code and observe your model saved under "trained_models" folder

To test model;

  1. Put your test images under "test_images" folder
  2. In "test_model.py" code, insert parameters like your pretrained model name, test image name etc
  3. Run the code and observe console