Research on Disparity Map Generation Method of Underwater Target Based on the Improved SIFT Algorithm

Di Wang; Yong Jie Pang

doi:10.4028/www.scientific.net/AMM.741.701

Paper Titles

Information Security Protection Design of Electric Vehicles Charging Station
p.681

Inverse Kinematics Analysis of a 2-Dof Spherical Parallel Shift Manipulator
p.687

Kinematic Analysis of 4-UPU Parallel Mechanism Based on Influence Coefficient
p.691

Monocular Camera Calibration Method for Robot System
p.697

Research on Disparity Map Generation Method of Underwater Target Based on the Improved SIFT Algorithm
p.701

Simulation Analysis for the Walking Mechanism of the Underground Exploration Robot Based on MATLAB and ADAMS
p.705

Size-Sorting and Measurement System of Safety Belt Pin Based on Machine Vision
p.709

Study of the Two-Phase Operation Mechanism of Three-Phase Motors Based on LC Phase Shifting Method
p.714

Multivariable Decoupling Control Based on TC Control in the Diving and Floating Process of AUV
p.720

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vol. 741Research on Disparity Map Generation Method of...

Research on Disparity Map Generation Method of Underwater Target Based on the Improved SIFT Algorithm

Abstract:

In order to obtain the depth information of the underwater target, it’s necessary to generate the disparity map based on binocular vision stereo matching. In the circulation water channel, the stereo matching experiments with underwater target were carried out by using the BM algorithm, SGBM algorithms and SIFT algorithm respectively. Then the characteristics of the disparity maps were analyzed for the three kinds of stereo matching algorithms. Compared with the BM algorithm and SGBM algorithms, the SIFT algorithm has been proved to be more suitable for underwater stereo matching. In order to obtain more feature points of underwater image, it is necessary to improved SIFT algorithm parameter. Underwater image matching experiments were made to determine the principal curvature coefficient γ. The results illustrated that the improved γ is better than the original value for underwater disparity map generation.