ABSTRACT
Department of Textile Engineering, University of Guilan, Rasht, Iran; E-mail: AKHaghi@yahoo.com
Abstract ................................................................................................... 72 6.1 Introduction .................................................................................. 73 6.2 Nanofibrous Membrane ............................................................... 78 6.3 Application of Nanofibrous Membrane in Filtration ................... 80 6.4 Effective Parameter in Filtration Process..................................... 81 6.5 Methods for the Characterization of Porosity
and Pore Structure ........................................................................ 83 6.6 3D Reconstruction of Pore Network ............................................ 88 6.7 Classification of 3D Geometry Group ......................................... 91 6.8 Vanishing Points and Lines .......................................................... 99 6.9 Image Features ........................................................................... 101 6.10 Feature Detection ....................................................................... 102 6.11 Feature Matching ....................................................................... 121 6.12 Keypoint Matching .................................................................... 124 6.13 Image Matching ......................................................................... 134 6.14 Reconstruction from Two Calibrated Views .............................. 135 6.15 Robust Estimation ...................................................................... 148
6.16 Bundle Adjustment..................................................................... 151 6.17 Visualization............................................................................... 152 6.18 Epipolar Rectification ................................................................ 153 6.19 Dense Matching ......................................................................... 155 6.20 Triangulation .............................................................................. 156 6.21 Texture Mapping ........................................................................ 157 6.22 Better Method for 3D Reconstruction of
Nanoporous Membrane .............................................................. 158 6.23 Conclusions ................................................................................ 158 Keywords .............................................................................................. 159 References ............................................................................................. 159
ABSTRACT
Nanoporous membranes are an important class of nanomaterials that can be used in many applications, especially in micro and nanofiltration. Electrospun nanofibrous membranes have gained increasing attention due to the high porosity, large surface area per mass ratio along with small pore sizes, flexibility, and fine fiber diameter, and their production and application in development of filter media. Image analysis is a direct and accurate technique that can be used for characterization of porous media. This technique, due to its convenience in detecting individual pores in a porous media, has some advantages for pore measurement. The three-dimensional reconstruction of porous media, from the information obtained from a twodimensional analysis of photomicrographs, is a relatively new research area. This chapter provides a detailed review on relevant approach of 3D reconstruction from two views of single 2D image. The review concisely demonstrated that 3D reconstruction consists of three steps, which is equivalent to the estimation of a specific geometry group. These steps include: estimation of the epipolar geometry existing between the stereo image pair, estimation of the affine geometry, and also camera calibration. The advantage of this system is that the 2D images do not need to be calibrated in order to obtain a reconstruction. Results for both the camera calibration and reconstruction are presented to verify that it is possible to obtain a 3D model directly from features in the images.
