An efficient polygon marine data update approach based on features matching

Abstract

The sea area under the jurisdiction of China is approximately 300 million km⁁2, which is one-fifth of the land area of China. There are a plenty of marine resources in China. So, some ocean development strategies are proposed in the 13th Five-Year Plan and 18th Party Congress of China. Now, different sources of ocean mapping determine data's differences in classification, scale, spatial location, geometry, topology and attribute description. Those differences make it hard to update and regulate ocean data. After having an array of comprehensive investigations in geometric characteristics of marine polygon data, a multi-level vector polygon features matching model is proposed in this paper. This model is used to match marine data, which means to find out vector features which represent a same object. So, the efficiency of updating is boosted and the longitudinal consistency of marine data is guaranteed. The model is divided into three steps: First, extract key characteristics of complex polygon vector features; Second, describe dynamic geometric characteristics of those features; Third, match and compare the characteristics through an advanced model based on shape and distance. In the first step of the model-extract key characteristics, the Douglas-Peucker Algorithm is improved in this paper to simplify polygon features, which is aimed at extracting those features' geometric characteristics and reserving key characteristics in those geometric characteristics. Also, the parameters, which is called distance threshold-T, in this simplification algorithm is modified dynamically. After the simplification of polygon features, a geometric model is used in this paper to describe shape characteristics and distance characteristics of a polygon feature. When it comes to evaluate the shape similarity of two polygons through a tangent space model (shape distance model), the complexity of this two polygon features should be similar. On the contrary, if a segment of a polygon is much more complex than the others of it, the shape distance function of this model will be unstable, which will result to the reduce of matching efficiency. However, in this paper, after a deep research in complexity of polygon features, an efficient matching algorithm, which evolves from the shape distance model, is proposed. To sum up, in this paper, the distance and simplification of shape characteristics in shape similarity is combined to evaluate the similarity between two polygon features. The Douglas-Peucker Algorithm is developed to simplify the complex polygon features, which promotes the speed of matching. What's more, the matching accuracy is improved through a shape description function in this paper. Also, some marine data with the scale of 1:10000 and 1:50000 is used to examine the methods proposed in this paper. The examination proves the methods in this paper is efficient and has strong robustness.