本文關(guān)鍵詞：三維FMF的HFCM水聲數(shù)據(jù)分割　出處：《計(jì)算機(jī)應(yīng)用研究》2017年10期 　論文類型：期刊論文

  更多相關(guān)文章： 模糊中值濾波(FMF) 直方圖模糊C-均值(HFCM) 數(shù)據(jù)分割 最大熵閾值法 半高斯模糊隸屬度函數(shù)

【摘要】：針對(duì)三維水聲數(shù)據(jù)背景復(fù)雜、受噪聲干擾嚴(yán)重等特點(diǎn),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(jì)算出模糊閾值;再結(jié)合半高斯模糊隸屬度函數(shù)對(duì)水聲數(shù)據(jù)進(jìn)行模糊中值濾波;最后采用HFCM算法對(duì)濾波后的數(shù)據(jù)進(jìn)行分割。對(duì)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。
[Abstract]:According to the three-dimensional acoustic data by complex background, serious noise characteristics, proposes a combination of 3D FMF HFCM acoustic data segmentation algorithm to improve the accuracy and efficiency of acoustic data segmentation. The algorithm firstly selects the three-dimensional filtering window, calculate the fuzzy threshold using the maximum entropy threshold method; combined with the fuzzy membership of fuzzy median half Gauss filtering of the acoustic data function; finally uses the HFCM algorithm to segment the data after filtering. The segmentation results on two different groups of 3D acoustic data show that this algorithm can effectively reduce the noise, segmentation effect is better than the HFCM and HFCM filter equalization FMF segmentation algorithm, fuzzy C- means algorithm and much better than the traditional segmentation efficiency.

【作者單位】： 杭州電子科技大學(xué)通信工程學(xué)院模式識(shí)別與信息安全實(shí)驗(yàn)室;
【基金】：國(guó)家自然科學(xué)基金資助項(xiàng)目(60802047) 浙江省大學(xué)生科技創(chuàng)新活動(dòng)計(jì)劃(新苗人才計(jì)劃)項(xiàng)目(2015R407075)
【分類號(hào)】：TB56
【正文快照】： 0引言隨著科技的發(fā)展,人們不再局限于對(duì)陸地資源的探索,而是選擇深入探索研究占地球70%面積的海洋世界。水下環(huán)境和物體的探測(cè)需求促進(jìn)了水下聲納成像技術(shù)的發(fā)展,由于水下環(huán)境復(fù)雜多變,聲納收集的數(shù)據(jù)極易受斑點(diǎn)噪聲干擾[1],加之水下存在各種雜質(zhì),極大地影響了水下聲納成像的，

本文編號(hào)：1386671