Usage

Create from zero

python tools/createDS.py
python main_voc.py
python tools/createSplit.py

Read from an existing dataset

cp sample.bbox_label.txt [dataset dir]
# edit sample.bbox_label.txt
python main_voc.py [dataset dir]/sample.bbox_label.txt

---

Extension

1. Add multi-bboxes for one class in one image.
2. Add voc2007 convert tool.
3. Create voc-format directly.

---

BBox-Label-Tool modify by sequence labeling

A simple tool for labeling object bounding boxes in images, implemented with Python Tkinter.

Data Organization

LabelTool
|
|--main.py # source code for the tool
|
|--Labels/ # direcotry for the labeling results

Dependency

python 2.7 win 32bit PIL-1.1.7.win32-py2.7

Startup

$ python main.py

Usage

1. Input the image dictory address(only for .jpg, but you can change it in code), and click 'Load'. The images along with first read will be loaded.
2. To create a seq. object, select the obj. and to draw a new bounding box as the original tool. Note that every bounding box should link to an object, every object links to one bounding box at most.

• To cancel the bounding box while drawing, just press .
• To delete a existing bounding box, select it from the listbox, and click 'Delete' at the bottom of bounding box listbox.
• To delete a existing obejct, simply click 'Delete' at the bottom of obj index listbox. Only unlinked object can be delete.

3. After finishing one image, click 'Next' to advance. Likewise, click 'Prev' to reverse. Or, input the index and click 'Go' to navigate to an arbitrary image.

• The labeling result will be saved if and only if the 'Next' button is clicked.