Car logo extraction and recognition using image processing

Hi everyone, I want to extract the logo from any car then compare it with template. the logo can be in front or back.

I'm trying to extract the logo using edge detection then compare it with template.

im=imread('download.jfif');
imgray=rgb2gray(im);

medianFilteredImage = medfilt2(imgray, [3 3]);
noiseImage = (imgray == 0 | imgray == 255);
noiseFreeImage = imgray; % Initialize
noiseFreeImage(noiseImage) = medianFilteredImage(noiseImage); % Replace.

imshow(noiseFreeImage);

imgray = noiseFreeImage;
imbin=imbinarize(imgray);

figure, imshow(imbin)
im=edge(imgray,"canny");


%se = strel('disk',40);
%erode = imerode(im,se);
%figure
%imshow(erode)
Iprops=regionprops(im,"BoundingBox","Area","Image");
area=Iprops.Area;
count=numel(Iprops);
max=area;
boundingbox=Iprops.BoundingBox;
for i=1:count
    if max<Iprops(i).Area
        max=Iprops(i).Area;
        boundingbox=Iprops(i).BoundingBox;
    end

end
im=imcrop(imbin,boundingbox);
im=bwareaopen(~im,50);
im=imcomplement(im);

binary=bwlabel(binaryImage,8);
imshow(binary)

I'm new to matlab and image processing and i did not find anything that can help me

Answer 1

[Image Analyst]

No, that's not going to work. There are so many edges in that image, or any other images of cars, that hoping to find out which edge is the logo is doomed to failure.

In my opinion you'd be best off using deep learning (transfer learning) to train a network to classify/find logos. Look for a demo on the Mathworks site where they use alexnet and retrained it to find something different.

[Image Learner]

Unfortunately, I have to use Image Processing. so can I use thresholding or anything that can help?

[Image Analyst]

If you have the Computer Vision Toolbox, you can try it this way:

https://www.mathworks.com/help/vision/ug/object-detection-in-a-cluttered-scene-using-point-feature-matching.html