https://github.com/angusg/truth-and-crop

Application for ground-truthing semantic segmentation datasets in PyQt4/OpenCV.
https://github.com/angusg/truth-and-crop

crop labeled-data labelme machine-learning opencv pyqt4

Last synced: about 1 month ago
JSON representation

Application for ground-truthing semantic segmentation datasets in PyQt4/OpenCV.

• Host: GitHub
• URL: https://github.com/angusg/truth-and-crop
• Owner: AngusG
• License: gpl-3.0
• Created: 2017-01-16T22:24:10.000Z (over 8 years ago)
• Default Branch: master
• Last Pushed: 2017-08-15T17:54:01.000Z (almost 8 years ago)
• Last Synced: 2025-05-07T12:13:18.951Z (about 1 month ago)
• Topics: crop, labeled-data, labelme, machine-learning, opencv, pyqt4
• Language: C++
• Homepage:
• Size: 5.54 MB
• Stars: 11
• Watchers: 4
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# truth-and-crop

Convenient GUI application for quickly ground-truthing semantic segmentation datasets in Python/OpenCV. Original dataset used with the tool can be downloaded from: https://dataverse.scholarsportal.info/dataset.xhtml?persistentId=doi:10.5683/SP/NTUOK9

![sample](images/sample.png)

### Dependencies

+ `python 3.4`

+ `pyqt 4.x`

+ `opencv 3.x`

+ `numpy 1.11.x`

+ `colorama 0.3`

+ `natsort=5.0.x`

+ `scikit-image 0.12.x`

If using Anaconda, you can use the provided `environment.yml` file with `conda env create -f environment.yml`, which will create a virtual environment `tnc-py34`.

### Usage

```bash

source activate tnc-py34

python truth_and_crop.py

```

### Buttons

+ __Input File__ - Browse to image file to load.

+ __Output Path__ - Browse to root folder where output should be saved. Three subfolders are automatically created here.

+ __Previous/Next Image__ - If other images were found in same folder as Input File, you can jump between images with these buttons.

+ __Refresh__ - Discards changes.

+ __Crop__ - Switch between annotation mode and cropping mode.

+ __Toggle__ - Toggle annotations on and off to make it easier to see raw image. SLIC is only run on the image for the first toggle, subsequent toggles are much faster.

+ __Save__ - To write all cropped images and masks into appropriate subfolders under the path specified by 'Output Path'.