DEEP LEARNING BASED FEATURE MATCHING AND ITS APPLICATION IN IMAGE ORIENTATION

Matching images containing large viewpoint and viewing direction changes, resulting in large perspective differences, still is a very challenging problem. Affine shape estimation, orientation assignment and feature description algorithms based on detected hand crafted features have shown to be error prone. In this paper, affine shape estimation, orientation assignment and description of local features is achieved through deep learning. Those three modules are trained based on loss functions optimizing the matching performance of input patch pairs. The trained descriptors are first evaluated on the Brown dataset (Brown et al., 2011), a standard descriptor performance benchmark. The whole pipeline is then tested on images of small blocks acquired with an aerial penta camera, to compute image orientation. The results show that learned features perform significantly better than alternatives based on hand crafted features.

Affine Shape Comparison using Different Distances

In this work, we propose to compare affine shape using Hausdorff distance (HD), Dynamic Time Warping (DTW), Frechet (DF), and Earth Mover distance (EMD). Where there is only a change in resolution shape distance are computed between shape coordinates because the distance is not invariant under rotation or affinity. In case of transformation, distances are calculated not between shape coordinates but between Arc length or Affine Arc length. Arc length is invariant under rotation while Affine Arc length is invariant under affinity. The main advantage is invariance under change of resolution, rotation, and affinity.

Finite element quasi-interpolation and best approximation

This paper introduces a quasi-interpolation operator for scalar- and vector-valued finite element spaces constructed on affine, shape-regular meshes with some continuity across mesh interfaces. This operator gives optimal estimates of the best approximation error in any Lp-norm assuming regularity in the fractional Sobolev spaces Wr,p, where p ∈ [ 1,∞ ] and the smoothness index r can be arbitrarily close to zero. The operator is stable in L1, leaves the corresponding finite element space point-wise invariant, and can be modified to handle homogeneous boundary conditions. The theory is illustrated on H1-, H(curl)- and H(div)-conforming spaces.

Affine Shape Descriptor Algorithm for Image Matching and Object Recognition in a Digital Image/Algoritmo para clasificación de formas invariante a transformaciones afines para objetos rígidos en una imagen digital

The use of shape, as a mean to discriminate between object classes extracted from a digital image, is one of the major roles in machine vision. The use of shape has been studied extensively in recent decades, because the shape of the object holds enough information for its correct classification; additionally, the quantity of memory used to store a border is much less than that of the whole region within it. In this paper, a novel shape descriptor is proposed. The algorithm demonstrates that it has useful properties such as: invariance to affine transformations that are applied to the border (e.g., scales, skews, displacements and rotations), stability in the presence of noise, and good differentiability between different object classes. A comparative analysis is included to show the performance of our proposal with respect to the state of the art algorithms.

Mechanically programmed shape change in laminated elastomeric composites

Laminated elastomeric composites exhibit non-affine shape change following a simple, room temperature tensile deformation.