Segmentation Revisited

Posted: February 10, 2012 at 11:36 am

After spending a week trying to debug memory corruption error I found the problem, I was attempting to write outside of the bounds on the reconstruction image, because a number of percepts were the size of the entire frame. Once I put in a condition to ignore these large percepts (segmentation failures), I got reconstructions that look like this:

Obviously, this is incorrect. All previous reconstructions then have an artificial sense of cohesion because of this error. I have revisited the segmentation and made some further changes. Thus far I am still getting some large segments that span the whole image, but the subsequent reconstructions are a marked improvement:

Now I can return to merging percepts across frames and higher level clustering.