Image Processing

Prof. Daniel Keren, dkeren@cs.haifa.ac.il

EXAMPLE QUESTIONS (WITH NEW ONES)

PROJECTS

1) Restoring noisy color images. The goal will be to find methods which use the correlation between the different color channels, in order to obtain better results than those achieved by de-noising each channel separately.

2) Homomorphic filtering. Remove the effect of non-uniform illumination.

3) Blind deconvolution. To remove blur caused by an UNKNOWN kernel.

4) OCR: recognize the text in a scanned page.

5) OCR for equations: recognize the structure of simple equations in a scanned page.

6) Mosaicing: "stitch" together movie frames to obtain a large image.

7) Detection: identify instance of human faces in images.

8) Detection: identify palm (hand), like in the airport...

9) Any interesting ideas you may have.

CVONLINE:

MATLAB:

See http://cs.haifa.ac.il/hagit/courses/ip/homeworkGeneral.html.

Recommended book: "Digital Image Processing" by Rafael C. Gonzalez, Paul Wintz
Second Edition, 1987.

Some sections directly relevant to the course (these may vary between different
editions -- check the section's name):

2.4: Neighbors and connectivity
3.1-3.3: Fourier transform.
3.7: Hough transform (we will study this later in the course).
4.1-4.5: Enhancement.
7.1: Edge detection.
7.4: Segmentation.

Note:

This does not cover the entire course -- there is other material which appears in the slides but not in this book.

There is also a book of the "Open University" in Hebrew.