Study On The Ability And Technology Of Ocean Information Detection Based On Satellite Images

The deteriorating marine environment and the frequent marine disasters have brought great damage to our economy. Thus, monitoring the changes of the marine environment timely and effectively, guiding the improvement of the marine environment and minimizing the damage become more and more important. With the continuous development of remote sensing technology and the improvement of sensor performance, remote sensing technology has become the most advantageous method to monitor a wide range of marine quickly. Therefore, the application of marine remote sensing technology to monitor the marine environment is of great research and practical significance.Firstly, ocean consists of water in constant motion and its environment is quite different from the environment of land and atmosphere. Secondly, remote sensing has its own characteristics, which is subject to the great difference between the land and ocean spectrum, the weak spectrum information extracted from the low-reflectivity ocean background and signal flares and so on. In addition, the dynamic ocean phenomena require fast and continuous monitoring. All of the above make the practical application of ocean remote sensing more difficulty. With the analysis of a large number of images, this thesis has investigated the ability of satellite remote sensing in ocean monitoring in the Bohai Sea, also analyzed the image features with different environment background and seasonal changes, and then further discussed the feasibility of fast anomaly detection to satellite images.By comparing the capabilities of several common remote sensing satellites in ocean monitoring, this thesis has analyzed the feasibility of evaluating the ability of satellite remote sensing in ocean monitoring based on the MODIS data in the Bohai Sea. The raw images are MOD35 data of the Bohai Sea in 2008, which are used to count the imaging capabilities of satellite remote sensing covering the Bohai Sea and their relevant cloud cover rate. This thesis has also evaluated the capabilities of remote sensing images of the Bohai Sea. The statistic results show different monitoring effects of regular, seasonal and emergency and other elements of the Bohai Sea.According to the results, it has showed the feasibility of extracting seasonal features of the MODIS image in the Bohai Sea based on the long time series MODIS images. By analyzing the true color composite maps of MODIS 1B data, the relevant layered maps and the background data of the Bohai Sea, the image features of different environment background and seasonal changes in each Bohai region have been found. Due to the quite stable horizontal distribution of suspended matter over the Bohai Sea, this thesis has selected the typical images in each region of the Bohai Sea in different seasons and has also established the standard MODIS image library of the Bohai Sea.Owing to the complicated optical properties of Case 2 waters, it is more difficult to extract the weak spectrum information from the low-reflectivity ocean background only based on the objective information extraction algorithms. On the basis of standard images of the Bohai Sea, this thesis has implemented the change detection of the dynamic images of the Bohai Sea. Considering the impacts of background environment, noise and other factors, it has analyzed the causes of spectral changes and has detected the possible abnormal area, which can be used to guide fine interpretation of this area to realize the extraction of abnormal information in the Bohai Sea. Finally, this thesis, taking the detection of short term oil spill and a longer time sea ice as examples, carried out the experiment of fast anomaly detection of satellite images in the Bohai Sea to get its technical methods and succeeded in extracting the oil spill and monitoring the dynamic process of sea ice in the Bohai Sea, both of which show desirable results.