HW 6, EECS 203A

Problem 5.16, Digital Image Processing, 2nd edition by Gonzalez, Woods.

Nasser Abbasi, UCI. Fall 2004

Question

Consider a linear position-invariant image degradation system with impulse response MATH Suppose that the input to the system is an image consisting of a line of infinitesimal width located at $x=a$ and modeled by MATH where $\delta $ is the impulse. Assuming no noise, find the output image MATH

Solution

In general, MATH where MATH is the noise. Hence since MATH, we have MATH

Now, I can solve this using spatial domain (convolution), or solve in Fourier transform domain, then inverse transform to get MATH. I'll try the direct spatial approach:

MATH

where $M,N$ are the dimensions of the image MATH

Since MATH then only when $x=a$, that we get a non-zero value for the the output image. At all other values for $x,$ MATH

Hence

MATH

Since MATH then Substitute MATH we get

MATH

How to evaluate these sums?

Since we are told that the input image is of infinitesimal width, then this means we can consider the sum MATH to have only one point. i.e. MATH

so MATH

To continue, best I could do is to look up the tables for integrals, and use the results for

MATH where $\func{erf}$ is the error function defined as MATH

Hence, using these, we get

MATH

Since $M,a$ are constants, then MATH is a constant, call it $k$

MATH

Then MATH where $A$ new constant$=k/2$

MATH

Notice that $g$ is a function of $a$ as well, since constant $A$ value depends on $a$ via the equation given above for $k$

$\ $

I