|
In recent years , the use of remote sensing images for ship detection and monitoring of research and technological development in the field of ocean remote sensing given high importance as remote sensing data is one of the most important marine applications . When the massive remote sensing image data for ship detection and surveillance , in order to timely and effective detection and extraction of ship target information , the need to carry out automatic detection algorithm . In this thesis, the application of remote sensing images of the ocean as a backdrop, the focus on remote sensing image ship target detection, ship wake detection, feature extraction and other aspects of marine research . Research papers , mainly in remote sensing image ship target , wake detection method and its application research. Firstly, studied the four common characteristics of wake generation mechanism , obtained by simulation of the sea , Kelvin wake, narrow V- wake, wake turbulence and internal wave wake characteristics . Allows us to understand the nature and understanding of the generation mechanism of different trails and spatial distribution patterns , in order to fully understand these types of wake model foundation. This paper analyzes the ship detection in remote sensing image research status ; experimental analysis is given by two-parameter CFAR algorithm , K- distributed CFAR algorithm and based on local window CFAR algorithm K- distribution characteristics and scope . In Ship Wake Detection algorithms , the paper summarizes the detection method of remote sensing image wake conventional structure , and highlights two categories detection methods, a class detection method is based on the geometric characteristics of the wake , and the other detection methods are based on the wake region and the entire remote sensing image background in statistical distribution differences. In the list of existing major ship wake detection method, for RADON, HOUGH other conventional algorithm , this paper proposes a new wake detection methods. Based on the theoretical analysis with simulation and real remote sensing images of the new detection methods were tested , and the test results are given analysis and evaluation . Finally, based on remote sensing images on the characteristics of the target ship , the ship's parameters studied characteristics of three types of parameters ( geometry , geography parameters , motion parameters ) features , and gives a quantitative parameter calculation method . In ship detection , wake detection and feature extraction studies , based on the characteristic parameters to achieve the estimated ship .
|