英文摘要 |
"The ship model ink paint test shows the streamline on the model surface, which indicate the situation of the flu-id flow around the ship hull. The valuable information, however, can only be recorded by photos or videos currently. Those recorded analogic visual images provide few information and are incapable for mathematical analysis. In this study, we convert the ink streamline photos taken from ink paint test into digitalized data by means of image processing technology, that make us able to perform further analysis and application. First of all, a streamline identification program is developed, in which Gaussian filter is used for photo prepro-cessing, and HSV color value is applied to identify the black ink streamline from the background color of yellow. Because it needs twenty pictures along the longitudinal direction to compose the total ink streamline, a pair of round markers to connect two continuous photos are introduced to perform the precious linkage of each other. After the identification process, the ink streamlines is converted into digitalized data and is attached on the surface of ship model. Once the coordinates of the ink streamline joint together with the ship model, one can then use drawing software to do zoom in/out, rotation, copy, distribute and etc. easily, which means the information from ink paint test can be shared with multiple personal. In order to understand the effectiveness of the digitalized ink streamline on the mathematical analysis, the simu-lated streamlines calculation by using CFD software is carried out. After a serial of comparison and analysis, it shows that the digitalized ink streamline data is effective in research purpose." |