How to identify a black dot in an RGB Image programmatically and then select a region in rectangle shape nearby to the black dot present in Image?

Hi, I want to select a particular region shown in red line with reference to point in the image(shown in black dot). Please refer the attached image.

Kindly Suggest a method to do the same using MATLAB. I need to find out that black spot in the image programmatically and then select the region based on that point present in the image. The selected region can be in the shape of rectangle.

[Ananthu Jayan]

I have attached the image above for the purpose of testing. @Image Analyst Kindly help me out.

[Jan]

Flags are thought to inform the editors and admins about contents, which is not covered by the terms of use, e.g. due to rudeness. If you do not get a useful answer, this is usually caused by the lack of useful information. Then provide more details, and explain what you have tried so far.

[Jan]

@Ananthu Jayan: Again I have removed your flag. Please stop using flags to catch attraction of readers, because this is not the purpose of flags. Thanks.

I think, you do not get an answer, because this is not clear: "Can I programmatically find the position of the black spot". You did not mention, how the position of the black spot is defined. Then it is impossible to find it by an algorithm.

Explain, what you want to do. Which region do you want to select by which method?

[Guillaume]

With the test image provided, finding the location of the black dot is trivial. That begs the question: why aren't you able to do it yourself? What is the difficulty?

With regards to the region selection, what is the criteria for the size of that region? Is it a user input (again, trivial to do) or is it based on the size of the leaf? (a bit harder but not that hard) or something else?

Answer 1

[Utkarsh Deshmukh]

Try this

    img = imread('img.png');
    reference_point = [450, 450];

    region_height = 200;
    region_width = 400;

    cropped_region = img(reference_point(1): reference_point(1)+region_height, reference_point(2): reference_point(2)+region_width, :);

by changing the reference point, you can change the top left coordinate. By changing the region_eidth and region_height, you can change the size of the region that you want to crop

[Ananthu Jayan]

Am getting error in the line cropped_region while setting the reference points. It says:

"Index exceeds matrix dimensions.

Error in anscheck (line 11) cropped_region = img(reference_point(1): reference_point(1)+region_height, reference_point(2): reference_point(2)+region_width, :); "

[Jan]

This means that your image is smaller than the selected rectangle. Of course you have to adjust the values of Utkarsh Deshmukh's answer.

[Ananthu Jayan]

I have attached the image above for testing. Can I programmatically find the position of the black spot in the image and then select a region accordingly.