ComputerVision- Homework 2 Solved

Hand in a complete report, as detailed as possible, including:your introduction
implementation procedure
experimental results (of course you should also try your own images)
discussion (what difficulties you have met? how you resolve them?)
conclusion
work assignment plan between team members. through E3 (the dataset will be also there) • Tasks:Hybrid image
Image pyramid
Colorizing the Russian Empire
Overview
A hybrid image is the sum of a low-pass filtered version of the one image and a high-pass filtered version of a second image. There is a free parameter, which can be tuned for each image pair, which controls how much high frequency to remove from the first image and how much low frequency to leave in the second image. This is called the “cutoff frequency”.

Image refer to: http://cvcl.mit.edu/hybrid/OlivaTorralb_Hybrid_Siggraph06.pdf

Image pyramid
Overview
Low pass filter

  Ideal low pass filter  

  Gaussian low pass filter  

Image refer to: Digital Image Processing, 3rd Ed., Rafael C. Gonzalez and Richard E. Woods 130

 

High pass filter
  Ideal high pass filter  

  Gaussian high pass filter  

Image refer to: Digital Image Processing, 3rd Ed., Rafael C. Gonzalez and Richard E. Woods

Procedure
Multiply the input image by (-1)x+y to center the transform.
Compute Fourier transformation of input image, i.e. F(u,v).
Multiply F(u,v) by a filter function H(u,v).
Compute the inverse Fourier transformation of the result in (3).
Obtain the real part of the result in (4).
Multiply the result in (5) by (-1)x+y.
Overview
An image Pyramid is a collection of representations of an image.

Refer to: http://cs.haifa.ac.il/hagit/courses/ip/Lectures/Ip11_MultiscaleRepx4.pdf

 

Gaussian image pyramid
Procedure

Set the finest scale layer to the image


For each layer, going from next to finest to coarsest

Obtain this layer by smoothing the next finest layer with a

Gaussian, and then subsampling it.


Colorizing the Russian Empire
Overview
Goal: automatically produce a color image from the digitized ProkudinGorskii glass plate images with as few visual artifacts as possible. 

The glass plate images record three exposures of every scene onto a glass plate using a red, a green, and a blue filter. In order to do this, extract the three color channels from the glass plate, then place and align one above the other so that the combination forms a single RGB color image.

Assume that a simple x,y translation model is sufficient for proper alignment. However, the full-size glass plate images are very large, the alignment procedure will need to be fast and efficient