Rough surface classification using point statistics from photometric stereo

Rough surfaces can be classified by the point statistics of their derivativ e fields, estimated using photometric stereo. Such a scheme is proposed and found to be more accurate and robust than image-intensity-based classifica tion. It is particularly effective when applied to directional surfaces, ev en under rotation. The scheme is therefore robust and economic - suitable f or many applications and worthy of further investigation. (C) 2000 Elsevier Science B.V. All rights reserved.