I am curious on how much detail can OpenMV detect. I have these small 1 inch diameter disks with very small cutting teeth that have a curf point a specific direction. Would this be able to detect which direction the curf is pointing?
Depends on the algorithm and the situation.
Would you mind going in to more detail on your answer. I am very new to this and feel this could be a very economical solution to my problem.
dear friend this depends on your working distance , focal length, lighting.
The camera has specific resolution.
Send us a picture of your product and then we can discuss further.
Here is a photo of a disk…
Basically I need to verify if it will be able to tell which way the teeth are pointing
Please take a picture using the OpenMV Cam.
Anyway, in your process is the gear always in the same place? If so, you just zoom on the gear teeth and then it should be easy.
At this point I haven’t purchased a OpenMV cam, so I am unable to use a photo from it it. However the goal is to have each disk in the same spot to analyze.
I am on a fact finding mission to verify this will work before making the investment to build the machine I need to facilitate the placing the proper ordinate disks.
So far it sounds promising.
Just clarify this is not a gear, its a very small toothed saw blade, and I need to make sure the teeth are pointing a specific direction when painting and screen printing the direction indicators required for the final part.
Would I go about having OpenMV compare two stored images of the teeth pointing each direction to what it sees?
Or would I employ a different detecting method that OpenMV uses.
Yeah, so, from a zoomed out pic it would be hard. If you could zoom in on a blade or just in general the top of it there’s a find_line_segments() method that will give you line segments that make up the part. Then you’d just do some filtering and math and the lines returned to get your results.
I would focus and zoom to the teeths and with find line tool the job would be possible.
Tips:
Use global shutter module.
Use backlight illumination for your application.