Machine Vision
How do you track 3 Similar objects with Geometric Matching?
Hello experts!
So I've modified the example program that was included with Labview Vision 2018 as well as the 2018 version of Labview. I didn't have anyone answer my previous post so I figured I could attempt to find a solution to the problem. I looked into multiple template/object tracking instead of finding similar pattern matches. The issue is that these three points I have to track are identical. I've tried pattern matching using the same image; however, the boundary switches from one to another.
I've provided the illustration above to prove my point. The other boundary would overlap and switch constantly. Is there a way to make the boundary stay with the object once it tracks it?
I also have the VI below as well as the three template images I've made. All I want to ask is if there is any way to track these three basically similar objects and have the boundary stay with its original object during initial tracking.
Below is the image it could possibly track:
Any help would be appreciated! Thank you so much for taking the time to read my message!
Alright, so I've thought of a plan of attack to fix this problem but I am stuck on how to implement it so far. I was thinking of tracking the region at every iteration so it looks for 3 dots.
Furthermore, to avoid it bouncing the boundary around I could specify search areas within the layout or define it with ROI's. I was thinking of using a rectangle ROI or an image display with a rectangle tool to receive the information. I was thinking of using the property node on the image display and using the ROI information.
The plan is to use a case structure to make sure that matches do exist, then use that boundary around the object and separate it into three slices. Each of the rectangular boundaries would be converted to ROI through the ctrl+click with multiple boundaries and an array of multiple ROIs.
You have three dots as shown below.
Boundaries could be made as such in the images below.
I'm stuck in understanding how to make the ROI code so that when it finds the first template picture of all three dots it can search in the three places within the template for the other dots. This instance would be repeated through the while loop and I can use the centroid of each dot to record the position and calculate the velocity.
I would really appreciate any insight or help into this. Thank you so much.
