Ask Your Question

Revision history [back]

Newbie - how to classify images based on average amount of colour?

Hi

I have what I think is a fairly simple objective (but am prepared to stand corrected). I need to sample a video stream about once a second and make a call as to whether the feed is showing one of two parts of the process.

For one part the video feed will be mostly static and the image will be mostly monochrome (although not actually a monochrome image i.e. it'll be full RGB or whatever), the camera involved only shows colour at close range.

For the second part of the process the image will be full colour.

I appreciate it's a fairly general question but I'm new to computer vision so I don't really know what I should be looking into in order to start classifying the images so I'd appreciate any pointers? I'm really just looking for some guidance on tutorials or videos I could use to get started.

I've gotten the plumbing working i.e. I've a scaffold application up and running and capturing Mats, it's just what to do with them next?

Language is C# using the OpenCVSharp library if that's relevant but happy to be pointed towards examples in other languages.

Thanks