Brandon
can you be more specific about your data, and possibly show screenshots? "classification" for imagery and "classification" for lidar mean very different things, so an unclassified image vs. unclassified lidar would be quite different.
Further, lidar is not imagery, unless you mean the intensity values from the LAS files. Do you have LAS files, or TIFF files, or something else?
And if you mean "find burned grass and trees" that's quite different from "there used to be a building here, now it's burned to the ground". If this is the same project as your other posts seeking advice on analyzing before/after images, lidar will be MUCH more effective for identifying burned buildings since you will have a dramatic difference in the DSM (digital surface model) that can be derived from the lidar. Delineating the extents of burned grassland would more likely be successful using imagery, presuming the imagery was captured recently after the fire (although the lidar intensity image may be effective for this - I don't know that I've seen lidar for a burned area).
re: your other posts, it might also be helpful to know how much total area you're examining (acres/square miles, total # of image files, ANY sort of quantification). If it's the entire western USA, the recommended solution may be different than if it's 10% of Sonoma County CA.
Cody B.
