I just finished creating histograms for all three channels (i.e. RGB). The histograms I got so far looks very promising. For the blood cell images, the Red channel histogram always has a very clearly identifiable peak value.
Team, how about writing a more abstract Image class, especially allowing us to access the image data of the form image[i][j]. The IplImage thing provides the image data as one huge 1D array and kind of hard to read and understand. This may cost in a latter stage in the implementation phase. But, it's always some additional trouble and time. Pls comment on this.
I will not be able to work on the project on Monday and Tuesday.
- Omega
P.S. I don't see others posts in the blog. Please post your current progress to the blog
Subscribe to:
Post Comments (Atom)
1 comment:
First of all apologies for not updating the blog with what I have been doing so far. Several things got in the way, and now only I started the *real* work.
As for the question you raised about iplImage - well, it would be good to have something like the Matlab sort of structure so we could use the more convenient 2D structure you suggested. However, I am not sure how feasible this would be if we are going to reuse the CV functions.
If you feel like it, please go ahead and implement it. We could always find a use for such a thing.
Post a Comment