$1.$ Open the script next given In this script you have the main necessary commands, except for the highlighted with blue colour. Make sure you have the input images: sample.jpg$,$

$2.$ Convert the image in binary format, using im$2$bw$.$ You should find the appropriate threshold value, so that the binarisation operation applies to all objects from the input image. Include the results of the binarisation of one of the images and the value of the threshold value in your report.

$3.$ Find the connected components of your obtained binary images $($with bwlabel $)$ and visualise $($with label$2$rgb$)$ the detected connected components.You have this functionality in the provided script. You do not need to include this visualisation in your report.

$4.$ Compute different characteristics of the connected components $($with regionprops$).$ Visualise the bounding boxes of the objects $($BoundingBox field from the regionprops function output, rectangle$($Position$,)).$ You have this functionality in the provided script. It is not required to include this visualisation in your report.

$5.$ Develop a function to distinguish arrows from other objects. What does differ the arrows from the other objects? You may use any ideas from lectures, the previous lab session or common sense for your function. Include the brief description of main idea of your function in your report and the actual code of the function in the appendix of your report. Hint: arrows have points of different colour.

$6.$ Find the starting red arrow. You have this functionality in the provided script.

$7.$ Develop a function to find the label of the next nearest object to which the current arrow indicates.

Hint $1$: to set a line it is enough to have two points.

Hint $2$: for each arrow you may extract the centroid point and the yellow one.

Hint $3$: a vector $($x$2$ x$1,$ y$2$ y$1)$ points to the direction from the point $($x$1,$ y$1)$ to the point $($x$2,$ y$2).$

$8.$ Apply your functions to find the treasure in the images. Visualise the treasure and the path to it from the starting arrow. You have this functionality in the provided script. Include your visualisation in your report for all images.

$9.$ Other solutions of this task are possible. If you propose a different solution, include a brief description of it and provide a diagram of the main idea of your solution. Include your functions in your report.

close all;

clear all;

$\%\%$ Reading image

im $=$ imread$(\text{'}$Treasure$\_$simple.jpg$\text{'})$; $\%$ change name to process other images

imshow$($im$)$; pause;

$\%\%$ Binarisation

bin$\_$threshold $=0$; $\%$ parameter to vary

bin$\_$im $=$ im$2$bw$($im$,$ bin$\_$threshold$)$;

imshow$($bin$\_$im$)$; pause;

$\%\%$ Extracting connected components

con$\_$com $=$ bwlabel$($bin$\_$im$)$;

imshow$($label$2$rgb$($con$\_$com$))$; pause;

$\%\%$ Computing objects properties

props $=$ regionprops$($con$\_$com$)$;

$\%\%$ Drawing bounding boxes

n$\_$objects $=$ numel$($props$)$;

imshow$($im$)$;

hold on;

for object$\_$id $=1$ : n$\_$objects

rectangle$(\text{'}$Position$\text{'},$ props$($object$\_$id$).$BoundingBox, 'EdgeColor', $\text{'}$b$\text{'})$;

end

hold off; pause;

$\%\%$ Arrow$/$non$-$arrow determination

$\%$ You should develop a function arrow$\_$finder, which returns the IDs of the arror objects.

$\%$ IDs are from the connected component analysis order. You may use any parameters for your function.

arrow$\_$ind $=$ arrow$\_$finder$()$;

$\%\%$ Finding red arrow

n$\_$arrows $=$ numel$($arrow$\_$ind$)$;

start$\_$arrow$\_$id $=0$;

$\%$ check each arrow until find the red one

for arrow$\_$num $=1$ : n$\_$arrows

object$\_$id $=$ arrow$\_$ind$($arrow$\_$num$)$; $\%$ determine the arrow id

$\%$ extract colour of the centroid point of the current arrow

centroid$\_$colour $=$ im$($round$($props$($object$\_$id$).$Centroid$(2)),$ round$($props$($object$\_$id$).$Centroid$(1)),$ :$)$;

if centroid$\_$colour$($:$,$ :$,1)>240$ && centroid$\_$colour$($:$,$ :$,2)<10$ && centroid$\_$colour$($:$,$ :$,3)<10$

$\%$ the centroid point is red, memorise its id and break the loop

start$\_$arrow$\_$id $=$ object$\_$id;

break;

end

end

$\%\%$ Hunting

cur$\_$object $=$ start$\_$arrow$\_$id; $\%$ start from the red arrow

path $=$ cur$\_$object;

$\%$ while the current object is an arrow, continue to search

while ismember$($cur$\_$object, arrow$\_$ind$)$

$\%$ You should develop a function next$\_$object$\_$finder, which returns

$\%$ the ID of the nearest object, which is pointed at by the current

$\%$ arrow. You may use any other parameters for your function.

cur$\_$object $=$ next$\_$object$\_$finder$($cur$\_$object$)$;

path$($end $+1)=$ cur$\_$object;

end

$\%\%$ visualisation of the path

imshow$($im$)$;

hold on;

for path$\_$element $=1$ : numel$($path$)-1$

object$\_$id $=$ path$($path$\_$element$)$; $\%$ determine the object id

rectangle$(\text{'}$Position$\text{'},$ props$($object$\_$id$).$BoundingBox, 'EdgeColor', $\text{'}$y$\text{'})$;

str $=$ num$2$str$($path$\_$element$)$;

text$($props$($object$\_$id$).$BoundingBox$(1),$ props$($object$\_$id$).$BoundingBox$(2),$ str$,$ 'Color', $\text{'}$r$\text{'},$ 'FontWeight', 'bold', 'FontSize', $14)$;

end

$\%$ visualisation of the treasure

treasure$\_$id $=$ path$($end$)$;

rectangle$(\text{'}$Position$\text{'},$ props$($treasure$\_$id$).$BoundingBox, 'EdgeColor', $\text{'}$g$\text{'})$;