基于水下地形图像匹配的辅助导航技术研究

发布时间：2018-09-08 13:22

【摘要】：目前,水下潜器采用的导航系统大多是以惯性导航为核心进行作业任务,然而随着时间的积累,惯性导航的定位误差会越来越大,所以长时间的航行中,惯导所提供的导航定位信息也会不够精确,影响作业效率。而水下辅助导航技术可以有效地修正惯导的导航误差,并且具有定位精度高、时效强、隐蔽性好及自主性程度高等优点。利用水下地形高程信息的水下辅助导航系统凭借低廉的价格、水下设备性能水平的提高、以及对惯性导航误差的有效修正使水下潜器的工程适用性大大增强。因此,本文将利用水下潜器搭载的多波束测深系统实测水下地形高程数据,将其转换成声纳图像,基于图像处理算法,研究一套基于水下地形的辅助导航方法。本文首先将由多波束测深系统得到的数据进行插值处理,转换成声纳灰度图像,提取水下地形图像的纹理特征、不变矩特征,并分别分析各特征在图像加噪、旋转、改变分辨率、改变探测点等不同条件下,特征参数值所受的影响,选取抗干扰能力强的特征参数构建圆形模板下特征向量。其次,对不同种类水下地形适配性问题进行分析,设计依据水下地形图像特征参数的适配性判断方法,并与传统适配性判断条件对比,讨论两种方法优缺点,归纳总结出新的水下地形适配性判断条件。再次,为保证水下潜器定位导航精度,提高导航效率,本文设计基于水下地形图像灰度的粗匹配、结合基于水下地形图像特征向量的精匹配的分层匹配方式,并与传统TERCOM地形匹配方法对比分析,阐明基于水下地形图像匹配的优点;对比分析最大互相关法计算航向角,本文设计多子图坐标法航向角估计,减少计算时间,同时进一步提高定位精度。最后,为方便对算法高效、良好地使用,本文设计开发水下地形图像辅助导航界面,阐述界面模块功能及操作方法。
[Abstract]:At present, the navigation system of underwater vehicle is mostly based on inertial navigation. However, with the accumulation of time, the positioning error of inertial navigation will be more and more serious, so in the long time navigation, The navigation and positioning information provided by inertial navigation will not be accurate enough, which will affect the efficiency of operation. The underwater aided navigation technology can effectively correct the navigation error of inertial navigation, and has the advantages of high positioning accuracy, strong time, good concealment and high degree of autonomy. The underwater aided navigation system based on the elevation information of the underwater terrain greatly enhances the engineering applicability of underwater submersible because of its low price, the improvement of underwater equipment performance level and the effective correction of inertial navigation error. Therefore, in this paper, the elevation data of underwater terrain measured by multi-beam bathymetric system carried by underwater submersible are converted into sonar images, and a set of aided navigation method based on underwater terrain is studied based on image processing algorithm. In this paper, the data obtained from the multi-beam bathymetric system are firstly interpolated and converted into sonar grayscale images, then the texture features and moment invariant features of underwater terrain images are extracted, and the noise and rotation of each feature in the image are analyzed, respectively. Under different conditions, such as changing the resolution, changing the detection point, the characteristic parameter value is affected, and the characteristic parameters with strong anti-interference ability are selected to construct the characteristic vector under the circular template. Secondly, the adaptability of different kinds of underwater terrain is analyzed, and the adaptability judgment method based on the characteristic parameters of underwater terrain image is designed, and the advantages and disadvantages of the two methods are discussed by comparing with the traditional adaptability judgment conditions. A new judgment condition for the adaptability of underwater terrain is summarized. Thirdly, in order to ensure the positioning and navigation accuracy and improve the navigation efficiency of underwater submersible, this paper designs a grayscale coarse matching based on underwater terrain image and a hierarchical matching method based on fine matching of underwater terrain image feature vector. Comparing with the traditional TERCOM terrain matching method, the advantages of underwater terrain image matching are illustrated, and the maximum cross-correlation method is compared to calculate the heading angle. At the same time, the positioning accuracy is further improved. Finally, in order to facilitate the efficient and good use of the algorithm, this paper designs and develops the underwater terrain image aided navigation interface, and describes the function and operation method of the interface module.
【学位授予单位】：哈尔滨工程大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41

【参考文献】