The Software Design Of High-accuracy Displacement Measure System Based On FPGA

Nowadays, meteorological facsimile maps are widely used, especially in the field of marine shipping. By receiving meteorological facsimile maps, navigation personnel can know the weather and sea condition in time. In the Integrated Navigation System, information from different navigation devices is reunited to be displayed on one synthetic platform. Meteorological facsimile maps are broadcast in dot matrix format, scholars have put forward many algorithms to get useful information (isobars, isotherms, etc.) on the meteorological facsimile maps, then turn the dot matrix data format to vector data format. By doing so, the interested information can be added to the electronic chart, making it easier for users to be aware of the weather and sea conditions on the route.In this thesis, the key research in vectorizing meteorological facsimile maps-vectorization of isobars and extraction and recognition of annotations are studied. According to the characteristics of meteorological facsimile maps, the methods of separation of graphic and text can be used to eliminate the base map such as latitude and longitude lines and coastlines. Then on the basis of analyzing theories and methods of character recognition and image vectorization, algorithms are proposed to extract annotations and vectorize isobars. The main research works of this thesis are as followed:(1) An annotation extraction algorithm based on the orientation of pressure annotations is studied.(2) The annotation characters recognition is implemented by template matching.(3) An isobars thinning algorithm based on tracking method is proposed to turn the isobars from multi-pixel to single pixel.(4) An isobars extraction algorithm based on α-β filtering technique is proposed for removing noise and connecting broken isobars simultaneously, getting the seamless and smooth isobars.(5) A vectorization algorithm of polygonal approaching is used to compress the data of isobars.Experimental results show that after the removal of base map, by algorithms mentioned above, annotations on the meteorological facsimile map can be extracted and recognized and isobars can be vectorized successfully.