This will create a $144\backslash$times $513$ sinogram array called p in which each row is a projection through an

object at a specific angle, $\backslash$phi $.$ The $513$th element of each row is always zero$-$valued. The extra element

in each row was added so that the number of columns is odd for the imrotate command, which I

used to generate the set of projections. The origin of each projection is the middle $(257$th$)$ element.

This corresponds to projection lines through the $(257,257)$ element in the original $513\backslash$times $513$ object

array used to generate the projections. Each row represents a different angle in $1\mathrm{.}25$o increments,

starting at $0$o and continuing in the counterclockwise direction up to $178\mathrm{.}75$o $.$ In the original image

notation, the positive sense of the y $($vertical$)$ direction is the top of the image. Note this is opposite

of the usual sense for indexing arrays in Matlab! When using FFT or IFFT commands $($including

the $2-$D versions$)$ make sure that you use array sizes that are powers of $2.$ Because the $513$th column

is all zeros in p$,$ you can do this by truncating the $513$th element in the x$-$domain. If you are not

using FFT or IFFT $($for example when using convolution filtered backprojection$)$ you do not need

to truncate the rows to $512$ elements.

In all cases, include both quantitative and qualitative assessments of the effects observed. If you

show me a correctly reconstructed image in person or via email $($only email jpg format of image$),$

I can email you the $$perfect$$ image array I used to generate the projections if you would like a

ground$-$truth image for numerical comparison. do not use iradon or radon for matlabcodeComparison of Resolution of Data for Filtered Backprojection: Write

Matlab programs to perform filtered backprojection using either the Ram$-$Lak or

Shepp$-$Logan filters and performing filtering in either the ${}_x^{\text{'}}-$domain or in the $x^{\text{'}}-$

domain. Experiment with changes in the spacing of the angles used $($for example use

only odd numbered rows of data$)$ and experiment with the $x^{\text{'}}$ sample spacing $($for

example use only odd numbered elements in each row$).$ Note that if you eliminate

the column containing the origins of the projections $(513$th in the original projection

data array$),$ you will have problems with the reconstruction. Discuss the effects of

lowering the data resolution in each of these two ways. do not use radon or iradon in matlab code.