Implemented in python. Dependencies: numpy, opencv, sklearn.
-
Simple pixelwise skin detection using classification tree.
-
Histogram backprojection.
Data and tools:
-
Skin Segmentation Data Set from https://archive.ics.uci.edu/ml/datasets/Skin+Segmentation
-
face.png image from http://graphics.cs.msu.ru/ru/node/899
-
3D projection of skin pixels was done by 3D Color Inspector/Color Histogram Tool http://imagej.net/plugins/color-inspector.html
Some results:
TODO:
1. GMM http://www.ghvandoorn.nl/skindetection.html
2. Make demo with user input.
3. Evaluate methods on additional datasets.
4. Try other color spaces http://pesona.mmu.edu.my/~johnsee/research/papers/files/rgbhcbcr_m2usic06.pdf
5. Incorporate skin detection in face detection framework like https://github.com/collinhover/face_detector
6. Add video processing example.
7. Maybe adopt grabcut to this task?
DATA:
http://lbmedia.ece.ucsb.edu/research/skin/skin.htm
http://www.feeval.org/Data-sets/Skin_Colors.html
http://web.fsktm.um.edu.my/~cschan/project1.htm
https://github.com/Maxfashko/CamVid
PAPERS:
"A Survey on Pixel-Based Skin Color Detection Techniques"
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.5.521&rep=rep1&type=pdf
"Statistical Color Models with Application to Skin Detection"
http://www.hpl.hp.com/techreports/Compaq-DEC/CRL-98-11.pdf
http://www.cse.psu.edu/~rtc12/CSE586Spring2010/samplePresentation.pdf
"Lecture 5: Face Localization and Skin Color Detector"
http://alumni.media.mit.edu/~maov/classes/comp_photo_vision08f/lect/05_skin_detection.pdf
"Adaptive Learning of an Accurate Skin-Color Model"
http://lbmedia.ece.ucsb.edu/resources/ref/fg04ppt.pdf