Hi Madhubalan,
The target feature seems to refer to the detectable object/feature, which in this case could be a smaller image of the printer alone. You would then run the corresponding image operations as for the whole image and produce the histogram for the comparison.
Anyway, I see you're still working on the method from the "Fast Object Detection Based on Color Histograms and Local Binary Patterns" paper. Earlier in another one of your posts we looked at the LBP part but apparently you would like to reproduce the whole method. By no means do I now intend to disrespect you and I don't, of course, see anything wrong in your idea but maybe I can give you tips to get more/better results in your search for support.
People in these forums, and in general, are much more eager to put some effort to help you when you first show the work that you yourself have put into the whole problem. It is a much more pleasant interaction when it has more the feel of cooperation or a shared challenge instead of, well, slave work, to be extreme. Most of us don't do this for work but as a voluntary contribution for the community.
So, if possible, please share us details about your project, about your LabVIEW skills and experience, what you have already managed to put together and what type of application are you ultimately going for. I'm sure that everybody would consider more to help you out with ideas or even bits of code.
.JPG){kind=link}
