【摘要】：水聲數(shù)據(jù)可視化的研究與應(yīng)用為海底地貌觀測,海洋資源勘探,水下救險以及海洋軍事等方面提供了有力的技術(shù)保障。本文針對水聲數(shù)據(jù)可視化中的數(shù)據(jù)預(yù)處理以及體繪制中的傳輸函數(shù)作了深入地研究,主要研究內(nèi)容和成果如下:首先,針對三維水聲數(shù)據(jù)背景復(fù)雜、受噪聲干擾嚴(yán)重等特點,提出一種結(jié)合三維FMF的HFCM水聲數(shù)據(jù)分割算法,以提高水聲數(shù)據(jù)分割的精度和效率。該算法首先選取三維濾波窗口,利用最大熵閾值法計算出模糊閾值;再結(jié)合半高斯模糊隸屬度函數(shù)對水聲數(shù)據(jù)進(jìn)行模糊中值濾波;最后采用HFCM算法對濾波后的數(shù)據(jù)進(jìn)行分割。對兩組不同的三維水聲數(shù)據(jù)進(jìn)行分割處理的結(jié)果表明該算法能夠有效地降低噪聲干擾,分割效果要優(yōu)于未濾波的HFCM以及均衡FMF的HFCM分割算法,并且在分割效率上要明顯優(yōu)于傳統(tǒng)的模糊C均值算法。其次,為了獲得良好的水聲數(shù)據(jù)三維顯示效果,設(shè)計不透明度傳輸函數(shù)和顏色傳輸函數(shù)。首先將體數(shù)據(jù)的標(biāo)量值映射為不透明度值;再計算體數(shù)據(jù)的梯度,根據(jù)所有體數(shù)據(jù)點的梯度幅值設(shè)計不透明度傳輸函數(shù),最終設(shè)計出基于標(biāo)量值和梯度幅值的二維不透明度傳輸函數(shù);然后設(shè)計了把標(biāo)量值、梯度幅值以及二階導(dǎo)數(shù)映射到HSV彩色空間的顏色傳輸函數(shù)。將設(shè)計的傳輸函數(shù)應(yīng)用到體繪制中,相比于標(biāo)量值不透明度傳輸函數(shù),引入梯度的二維不透明度傳輸函數(shù)的繪制結(jié)果強(qiáng)化了邊界信息,突出了目標(biāo)。顏色傳輸函數(shù)的設(shè)計使得水下物體更加容易辨識。本文首先對水聲數(shù)據(jù)進(jìn)行降噪分割處理,然后設(shè)計不透明度和顏色傳輸函數(shù),利用光線投射算法對處理后的水聲數(shù)據(jù)進(jìn)行體繪制。實驗結(jié)果表明,本文的研究工作能夠獲得良好的水下環(huán)境顯示效果,不僅突出了目標(biāo)物,而且繪制出的圖像也十分清晰,更加有利于后續(xù)工作的觀察和分析。
[Abstract]:The research and application of the visualization of underwater acoustic data provide a powerful technical guarantee for the observation of the submarine topography, the exploration of marine resources, the underwater rescue and the marine military. In this paper, the data pre-processing in the visualization of underwater acoustic data and the transfer function in volume rendering are studied in-depth. The main research contents and results are as follows: First, the background of the three-dimensional underwater acoustic data is complicated, the noise is seriously disturbed, etc. An HFCM underwater acoustic data segmentation algorithm combined with three-dimensional FMF is proposed in order to improve the accuracy and efficiency of underwater acoustic data segmentation. The method comprises the following steps of: firstly, selecting a three-dimensional filtering window, calculating a fuzzy threshold by using a maximum entropy threshold method, and performing fuzzy median filtering on the underwater acoustic data in combination with a semi-Gaussian fuzzy membership function; and finally, dividing the filtered data by using an HFCM algorithm. The result of dividing the two groups of different three-dimensional underwater acoustic data shows that the algorithm can effectively reduce the noise interference, and the segmentation effect is better than the non-filtered HFCM and the HFCM segmentation algorithm of the balanced FMF, and the segmentation efficiency is obviously better than the traditional fuzzy C-means algorithm. Secondly, in order to obtain good underwater acoustic data three-dimensional display effect, the opacity transfer function and the color transfer function are designed. firstly, the scalar value of the volume data is mapped into an opacity value; the gradient of the body data is calculated, the non-transparent transmission function is designed according to the gradient amplitude of all the body data points, the two-dimensional opacity transmission function based on the scalar value and the gradient amplitude is finally designed, and the scalar value is designed, The gradient magnitude and the second derivative are mapped to the color transfer function of the HSV color space. The design transfer function is applied to volume rendering. Compared with the scalar value non-transparency transfer function, the drawing result of the two-dimensional opacity transfer function of the gradient is reinforced with the boundary information, and the target is highlighted. The design of the color transfer function makes it easier for the underwater object to be recognized. In this paper, the underwater acoustic data is subjected to noise reduction and segmentation, then the opacity and the color transfer function are designed, and the processed underwater acoustic data is subjected to volume rendering by using the light projection algorithm. The experimental results show that the research work of this paper can get good underwater environment display effect, not only has the target object been highlighted, but also the drawn image is very clear, which is more beneficial to the observation and analysis of the follow-up work.
【學(xué)位授予單位】：杭州電子科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TB56

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 高真;葉學(xué)義;周天琪;宋倩倩;;基于反距離加權(quán)插值的水聲數(shù)據(jù)可視化算法[J];計算機(jī)工程;2015年09期

2 卞紅雨;沈鄭燕;張志剛;劉翠;;基于雁群優(yōu)化的聲吶圖像快速閾值分割方法[J];聲學(xué)與電子工程;2011年03期

3 楊海亮;張森;唐勁松;;寬測繪帶多陣合成孔徑聲納成像的仿真研究[J];系統(tǒng)仿真學(xué)報;2011年07期

4 王士龍;徐玉如;龐永杰;;一種基于PSO優(yōu)化HWFCM的快速水下圖像分割算法(英文)[J];Journal of Marine Science and Application;2011年01期

5 宋坤坡;夏順仁;徐清;;考慮小波系數(shù)相關(guān)性的超聲圖像降噪算法[J];浙江大學(xué)學(xué)報(工學(xué)版);2010年11期

6 桑恩方;沈鄭燕;卞紅雨;葛光濤;;形態(tài)小波域聲納圖像去噪算法[J];數(shù)據(jù)采集與處理;2010年03期

7 周曉明;馬秋禾;肖蓉;;基于Canny算子的改進(jìn)的圖像邊緣檢測方法[J];影像技術(shù);2008年04期

8 嚴(yán)國萍;戴若愚;潘晴;劉袁緣;;基于LOG算子的自適應(yīng)圖像邊緣檢測方法[J];華中科技大學(xué)學(xué)報(自然科學(xué)版);2008年03期

9 郝穎明,朱楓;2維Otsu自適應(yīng)閾值的快速算法[J];中國圖象圖形學(xué)報;2005年04期

10 陳為;硬件加速反走樣體Splatting算法[J];計算機(jī)輔助設(shè)計與圖形學(xué)學(xué)報;2005年04期

相關(guān)博士學(xué)位論文 前1條

1 么彬;多子陣波束域高分辨水聲成像技術(shù)研究[D];哈爾濱工程大學(xué);2009年

相關(guān)碩士學(xué)位論文 前4條

1 劉加銀;海底聲學(xué)探測數(shù)據(jù)綜合可視化[D];國家海洋局第一海洋研究所;2015年

2 康健超;基于GPU的器官體數(shù)據(jù)光線投射算法研究[D];西北大學(xué);2011年

3 程鵬;圖像聲納發(fā)射與控制模塊的設(shè)計與實現(xiàn)[D];哈爾濱工程大學(xué);2010年

4 宋坤坡;水聲三維數(shù)據(jù)場的可視化技術(shù)研究[D];浙江大學(xué);2010年

，

本文編號：2503145