How do I measure the area of objects in a binary image, to be passed into the function bwareafilt?

So i have a binary image below:

and what I want to do is to extract the two smaller rectangle-like object from the binary image using bwareafilt(). However, I'm required to input a certain range as required for bwareafilt() and i would like to know how to calculate the area of each object in the binary image before specifiying a range for bwareafilt(). In other words, I want to know if there is any built in function that returns the area of each object in the image. For now, i had to experiment with different range via hardcode where:

bw3 = bwareafilt(binimg, [100 30000]);

The range value of 100 to 30000 allows me to extract the two rectangles from the image, without the large one above. I'm looking for a method where i wouldn't have to exhaustively type in ranges, hoping to extract the wanted object from the image.

Answer 1

[Walter Roberson]

regionprops() with the 'Area' option.

However, there is another approach:

bwareafilt(binimg, 2, 'smallest')

If what you want to do is get rid of the 1 largest when you do not know how many blobs there are, then there are several approaches:

• use bwlabel() or bwconncomp() or regionprops() to find out how many blobs there are, and then pass 1 less as the number to keep with 'smallest'
• or you could do binimg - bwareafilt(binimg, 1, 'largest')

[Stewart Tan]

@Walter Roberson thank you! regionprops() does the job!

[darova]

Please accept the answer

Walter will lose his reputation without your help