I am mapping mangroves over a period of several years using satellite imagery. I use QGIS semi automatic classification plug in. Supervised classification (maximum likelihood) as I can visually identify the trees. For one year I have only the panchromatic band available. The problem is that the spectral difference of the mangroves not only to other vegetation but to everything except water does not seem to be enough to differentiate them. I tried using a lot of training sites but still no success. It would be great to have some advice/ideas as to how to approach the task.
You could try an image segmentation approach but, I would not hold my breath on usable results. As far as application of a classification algorithm to panchromatic imagery, it is quite doubtful that you will get usable results because of the lack of any spectral separability associated with your target vegetation class. The only relevant information that you can hope to extract would be based on pattern, which is why a segmentation approach may be your best bet.
Given that you are attempting to extract mangroves, you really need to utilize multispectral data because pattern alone will not be an adequate indicator of the the target class. This is why it is critical to match data to the question before starting an analysis. In this case there is just no way of avoiding the reality that panchromatic imagery is not adequate for the intended classification.