Home

Awesome

Interactive-Segmentation-on-RGBD-Images-via-Cue-Selection

Python implementation of paper "Interactive Segmentation on RGBD Images via Cue Selection", CVPR 2016.

top

Dependencies

Python3.x (Tested with 3.5)
opencv
PyQt5
scikit-learn
PyMaxflow
Dijkstar

Installation

To use this code, please do:

  1. Clone the repo:

    git clone https://github.com/ZVsion/rgbd_image_segmentation
cd rgbd_image_segmentation

  2. To try the demo, please run:

    python main.py

If installed correctly, the GUI should look like this:

<p align="center"><img src="doc/gui1.png" align="center" width=450 height=350/></p>

You can add some foreground and background seeds in image, and then click Segment Image. After a few seconds, the result should look like this:

<p align="center"><img src="doc/segment_cues.jpg" align="center" width=350 height=450/></p>

Each color's label is presented as follow:

label

For more details about this algorithm, please see in paper and code.

Noted that we do not implement Grabcut postprocessing here.