Need to detect the edges of a rectangle using Image processing?

So I first need to read the image file of student ID card get the edges of the ID card and draw a rectangle covering the edges and the crop the detected edges of the image.

These are the methods I used so far

--> To read the image file - imread()

--> To convert to grayscale - rgb2gray()

--> To detect the edges - edge() using the Canny method

--> To get only edge data - bwconvhull()

--> To draw the rectangle over the ID - regionprops

These are my main idea the solutions works but its not that accurate. Pls if somebody knows it that would be really helpful

Answer 1

[drummer]

What if you use dilation-erosion approaches?

[Jeen Xavier]

This is my existing code which I have made its not the best, Do you know any suitable fix for my code to be more optimal?

f = imread('image-001.bmp'); %read the image file 
% Convert the image to a gray scale image
rgb = rgb2gray(f);

BW1 = edge(rgb,'Canny'); %putting the edge function to mask the image 
CH = bwconvhull(BW1);

[L n] = bwlabel(CH,8);
blobs = regionprops(L, 'BoundingBox');
% blobs(1).BoundingBox % the third and fourth row shows the width and height 
width = blobs(1).BoundingBox(3)
height = blobs(1).BoundingBox(4)

imshow(f);
rectangle('Position',blobs(1).BoundingBox,'Edgecolor','g');
% cropping the new image with respect to the rectangle created 
new_crop = imcrop(f, [blobs(1).BoundingBox(1) blobs(1).BoundingBox(2) blobs(1).BoundingBox(3) blobs(1).BoundingBox(4)]);
imshow(new_crop)