Quaternion Based Highlight and Shadow Detection for 3D Surface Recognition

This paper proposes a Quaternion based highlight and shadow detection algorithm in four source photometric stereo. This uses the concept of edge behavior of specularities i.e highlights are formed on top of the surface around the edge areas and shadows are formed on bottom of the surface around bottom edge pixels and this quaternion based approach is developed based on the fact that color edges only exist in a non homogeneous region. The specularities are efficiently detected nearby edge pixels by combining color channels as quaternion than the earlier methods.