Write a script which imports the image $($cracked$\_$wall.jpg$)$ and filter the image:

recall that a RGB color image is saved in Matlab as a $(nm3)$ matrix, where the red, green and blue channels are saved as

$(n$ x m x $1),(n$ x m x$2),(n$ x m x $3),$ respectively. im$\_$gray is the grayscale version of the image and it is saved as a $(n$ x m$)$

matrix. Plot the grayscale image im$\_$gray

calculate the im$\_$ratio as the ratio of the red channel over the mean of the three channels, i$.$e$.,($double$($im$($n$,$m$,1))+$

double$($im$($n$,$m$,2))+$ double$($im$($n$,$m$,3)))/3.$ Plot the newly calculated image im$\_$ratio. $($Convert the data type to double!$)$

Hint: the resulting image should have a single channel $($it will be a grayscale image$).$ You may also have to convert your uint$8()$

matrix into the datatype double$()$ before doing this.

create a mask im$\_$out to filter out parts of the image which are not red. The mask is a zero matrix $(0$ is black$)$ with the only non$-$

zero elements in correspondence of the red cracks. These elements are such that im$\_$ratio $>$ thresho$/$d $=1\mathrm{.}9.$ Create im$\_$out

that only show pixels with im$\_$ratio $>1\mathrm{.}9$ so that only the red cracks appear.

Hint: there will be other parts of the image that still shows. We call this noise and that is acceptable.