Brightness tool

OpenMV related project discussion.
Posts: 2
Joined: Thu May 25, 2017 3:52 am
Location: Netherlands

Brightness tool

Postby ikkez » Thu May 25, 2017 5:04 am

Hi there Forum members,

I’m an automation engineer from the Netherlands that bought some of the M4 models during the kickstarter campaign.
But until now there still in the package and I didn’t had any time to unpack the Camera’s.

We integrate PLC’s and machine vision into industrial applications and at the moment we have a project where we would like to test the newer faster M7 model, but reading the list of functions/Tools we can’t really find a tool for brightness detection.
Our project is quite simple, We have a black steel conveyor belt with yellow cookie dough transporting on there.

At the moment we have a camera from a different brand, where we have selected an area for brightness detection, when the cookie dough passes the Area an digital output is set. And from there we start measuring with an encoder on the conveyor to position the cookie dough.

The camera is now running at 100fps to get an accurate position. I noticed that the M7 should be able to run 85 FPS and I think it would be more than enough.
Is it possible to get a higher speed when we use only half of the image resolution or even set a smaller shutter time ? (I have no idea where the mentioned 85 is based on except for color and resolution)
Can we fulfill this job with an M7 ? and which tool could we use ?

Brightness tool.jpg
Brightness tool.jpg (36.62 KiB) Viewed 442 times
Something with 10 people and only 2 types of people that understand.
User avatar
Posts: 1325
Joined: Sun May 24, 2015 2:10 pm

Re: Brightness tool

Postby kwagyeman » Thu May 25, 2017 10:10 am

Please see the get statistics method in the image module under documentation. This method returns the mean, median, mode, stdev, lower quartile, upper quartile, min, and max of all the color channels in a specific ROI of the image. If you set the camera to grayscale mode then this should run at 85 FPS.

Um, so basically in your python script you'd call the get statistics method on an ROI where you expect each cookie to be at a certain time to measure the brightest.

Alternatively, even better would be to just call the find blobs function on the image using the inverse of whatever the color of the belt is. This will then tell you the ROI of all blobs in the image. Then you can just call the get statistics method on each blob location to measure the brightest. As long as your looking at a small resolution/window you should still be able to hit 85 fps.

Doing the above method saves you from having to figure out the location of each area where cookies should be.

Return to “Project Discussion”

Who is online

Users browsing this forum: No registered users and 1 guest