A bundled python application used to classify objects for a directory of images. This software MUST be run on Ubuntu 18.04.X. The object detection utilizes TensorFlow 1.12 and does not require users to install any packages.
Users can run the classify_directory executable located within the dist/classify_directory folder. The application looks in the folder original_images and labels each image. The labeled images are then saved with there corresponding bounding boxes in the folder classified_images.
Within the folder dist/classify_directory/model there are two files: A frozen inference graph and a label map. If you would like to use your own trained model, you should replace these files with your corresponding inference graph and label map.
This application is in beta stages so there are many updates that need to be made. A list of future updates are:
- Utilize a GUI to allow users to select their own directory
- Display labeled images within GUI