Autonomous system to build 3D model of underwater objects

Abstract

This thesis consists of introduction to 3-D Modeling and specifies the objectives of research. The background study to compare this work with the Human vision system and the stereo vision. Also it explains different steps involved in reconstruction of underwater objects. The steps being image acquisition, image enhancement, feature point detection, feature point matching, epipolar geometry, image rectification, 3D reconstruction, registration of all feature points and the final 3-D structure. The problem of 3-D reconstruction in underwater atmosphere is compared with the objects in air. Also this chapter gives the working of 3-D model reconstruction system and scope of the present work. Specifies major functionality of the work with different working modules like image acquisition, image enhancement, feature extraction, feature correspondence matching, Epipolar geometry, matrix estimation, metric reconstruction and camera calibration. Also the data of flow diagrams at level0, level1 and level2 are drawn. The specifies experimental setup and testing which explains laboratory arrangement for underwater object and analysis. The individual components of the system are tested to ensure that they are operated correctly. Unit tests exercise a small piece of functionality within a system. Typically, a test runs individual functions against known inputs, with the results verified against expected results. Unit tests not only ensure the test for successful outputs; but also test for appropriate error and boundary conditions. newline