I modified a great simple image annotator developed by Sebastian Perez to facilitate browsing forward and backward and to output data in per-image files in KITTI format for easy consumption for object detection using detectnet.
- Install Flask
$ pip install Flask
- cd into this directory after cloning the repo
- start the app
$ python app.py /images/directory
- you can also specify the directory you would like the annotations output to (image dir is the default)
$ python app.py /images/directory --out /labels
- specify port if you want (5555 is default)
$ python app.py /images/directory --out /labels --port 5556
- open http://127.0.0.1:5555/tagger in your browser
- only tested on Chrome
- This branch outputs data in KITTI format for easy consumption in DIGITS using detectnet
- filename is the same as the input image file prefix changed to .txt
- pertinent columns for detectnet are 0: label; 1: truncated; 4: xmin; 5: ymin; 6: xmax; 7: ymax
person 0.2 0 0.0 114.0 650.0 227.0 796.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
- draw a bounding box
- click on the image in the location of the first corner of the bounding box you would like to add
- click again for the second corner and the box will be drawn
- add a label for a box
- for the box you would like to give a label, find its id (noted in the top left corner of the box)
- find the label with the corresponding number
- enter the name you want in the input field
- press enter
- move to next image
- click the blue right arrow button at the top of the page (depending on the size of the image you may have to scroll down)
- move to the previous image
- click the left arrow button at the top of the page
- remove label click the red button on the label you would like to remove
- check generated data
- the output directory should contain one .txt file per image with KITTI format data