A Constrained Independent Component Analysis Based Photometric Stereo for 3D Human Face Reconstruction

In this paper, we propose an improved photometric stereo scheme based on the Lambertian reflectance model and the constrained independent component analysis (cICA) method. When we obtain an object's surface normal vector on each point of an image using the ICA model to reconstruct 3D shapes, we find that the x-axis, y-axis and z-axis values of the normal vector's coordinates are not arranged in turn. Thus, we use the cICA method to solve the problem. We obtain the correct normal vector's sequence form surface, and use a method for enforcing integrability to reconstruct 3D objects. We tested our algorithm on synthetically generated images for the reconstruction of object surfaces and on a number of real images captured from the Yale Face Database B. All the experimental results were compared to those of the existing photometric stereo approaches tested on the same images. The experimental results demonstrate that the proposed cICA method is better than some existing approaches.