Office Hours and Location: Monday 4:00 pm.
TA: Benoit Corda, e-mail: corda at cs.nyu.edu, office hours by appointment
Books: None
Course Structure: theory and laboratory work. Groups of three students will work together in developing vision applications. We will use the microsoft product Kinect as a plataform to acquire images, infrared depth maps, image segmentation, and skeletons.
See the following SDK
- Primesense SDK (windows, Linux, MacOs)
http://www.openni.org/downloadfiles/openni-compliant-middleware-binaries/33-latest-unstable
HowTo install on Linux:
http://www.keyboardmods.com/2010/12/howto-kinect-openninite-skeleton.html
- LibFreenect (no skeleton) for windows, Linux, MacOS
http://openkinect.org/wiki/Getting_Started
Projects will be developped in class and as homeworks. Evaluation is based on participation and project results.
Syllabus:
1. Introduction to Computer Vision.
Lecture 1.ppt (power point)
2. Format Images
There is a large library of computer vision algorithms including how to read and write image files
at
opencv
3. Image Measurements and Feature Detection
(Image pixels, oriented filters, derivatives)
Lecture 2.pdf (pdf document)
4. Image Segmentation and Graph Cut Method
(Lecture 3 is from Hiroshi Ishikawa, Ph.D. student at NYU and now professor
in Japan, see his page http://www.f.waseda.jp/hfs/index.html)
Lecture 3.pdf (pdf document)
This is a paper by Microsoft Cambridge Researchers with a more detailed development of Graph cut for Segmentation
Grabcut.pdf (pdf document)
Homework 1: Segmentation.
Deadline, February 28th.
Homework 1.pdf
Additional Materials
5. Shapes and Skeletons
Lecture 4.pdf
Homework 2: Camera calibration.
Additional Materials
Neural Networks
Lecture.pdf
Homework 3: face detection
Instructions