Trying to use cv.jit.blobs.orientation for a bridging app I’m making and I’m having a lot of difficulty understanding how to pull out the component area for more than one area. The abstraction cv.jit.blobs.orient.draw does this and renders it to a LCD but I need to instead separate out values as they will be sent on to another application. I presume each component has four coordinates for its bounding area but no matter how I look at or test the output matrix I rarely get multiples of four. Is anyone here familiar with the formatting of the output?
cv.jit.orientation does not provide the bounding box, only the orientation. That’s the reason why you have to plug in the output of cv.jit.moments into cv.jit.orient.draw, otherwise it wouldn’t know where the blobs are. If you dig deep into cv.jit.orient.draw, you’ll see that I’m simply using the blob area to determine the length of each line.
If you need a proper bounding box, use cv.jit.blobs.bounds.
If you want the values of these objects, use jit.iter. For all the cv.jit.blobs.* objects, there is always only one cell per blob.