Consideration of Various Noise Types and Illumination Effects for 3D Shape Recovery

Abstract

There are a variety of 3D Shape estimation methods. Broadly these methods can be classified into three types, namely, Contact, Transmissive and Reflective methods. The “Contact” method is generally based on some physical contact to acquire data while the “Transmissive” method is based on sending waves (like electromagnetic radiations, sound waves etc) through a body and recording data because of the interaction of wave particles with the object under consideration. The reflective model acquires data based on reflection of wave particles. The reflective method is broadly divided into optical and non-optical techniques. The optical methods under the reflective model can further be divided into “Passive” and “Active” Techniques. In active techniques, we are projecting light rays while in passive techniques; we simply capture the reflection of light rays without any projections. The passive methods can further be classified as Shape From X (Stereo, Motion, Shading, Focus etc). This chapter deals with Shape From Focus (SFF) which is a passive optical method. The objective of shape from focus is to find out the depth of every point of the object from the camera lens. Hence, we obtain a depth map which contains the depth of all points of the object from the camera lens where they are best focused. The aim of this chapter is to study the various factors (for example, different types of noise, illumination, window size) that affect SFF. It is shown that the illumination effects can directly result in incorrect estimation of depth map if proper window size is not selected during the computation of focus measure. The large window size results in blurring the image which gives the wrong impression of smoothness of the depth map. So it is important to find the optimum window size for accurate depth map estimation. Further, it is shown that the images need some kind of pre-processing to enhance the dark regions and shadows in the image. Additionally, a robust focus measure is also discussed in this chapter. This focus measure has shown robustness in the presence of noise as compared to the earlier focus measures. This new focus measure is based on an optical transfer function implemented in the Fourier domain. The focus measure is tested at various levels of noise, i.e., low, medium and high noise levels. The results of this focus measure have shown drastic improvement in estimation of depth map, with respect to the earlier focus measures, in the presence of various types of noise including Gaussian, Shot and Speckle noise.