Felenasoft

I'm curious how the size of the blue box is actually used by Xeoma. How is the calculation done to determine whether detected motion is greater than the "minimum size"?

I ask because I'm trying to detect humans walking through an area. However, humans have a tall, vertical rectangle shape when they move. The shape of the "minimum size of object" selector is a wide, horizontal rectangle.

So, what actually determines whether an object is big enough to be more than the minimum size? Is it, if the bounds of the minimum size rectangle are exceeded by the motion, then the motion is big enough?

That doesn't seem to be the case, since people are getting skipped by my current "minimum size" despite being much taller than the height of the minimum size box.

Here is an image that more clearly explains what I'm asking:

https://i.imgur.com/Q91vfkr.png

Here is another example of what I mean:



Blurred for privacy, but you can see the small pink box I draw in the middle is roughly how big the "minimum size of object" is. You can also see that I have the slider for "minimum size of object" set to just barely above the smallest possible setting.

The larger pink box on the left is a person walking by.

However, you can see at the top the yellow text says "Filtered out due to minimum object size settings". But, the person is obviously bigger than that tiny box in every way.

How is this working? How come the much bigger person walking by is being filtered out?

I'm interested too in knowing the answer, because I have tried several options and I did'nt found the way application does the trick...

Setting the minimum size of an object allows the system to ignore the movement of any object whose area is less than the area of minimum object.
The maximum object size is designed to block the signal when an object area exceeds an area of maximum object.

In the first example picture, the signal will not pass because the object area is less than area of the minimum object.