【摘要】：中國是一個(gè)水運(yùn)大國,內(nèi)河航道縱橫交錯(cuò),通江達(dá)海,水路交通十分發(fā)達(dá)。近幾年來,受國內(nèi)經(jīng)濟(jì)轉(zhuǎn)型、生產(chǎn)調(diào)整的影響,內(nèi)河船舶運(yùn)輸超載現(xiàn)象有大幅反彈的趨勢,給水上運(yùn)輸安全帶來了嚴(yán)重的危害。海事部門以船舶干舷值作為衡量船舶是否超載的主要標(biāo)準(zhǔn),因此船舶干舷值的測量及計(jì)算成為備受關(guān)注的研究熱點(diǎn)。本文提出了基于激光掃描的船舶干舷測量方案,具體研究內(nèi)容如下:船舶干舷值定義為船舶甲板到水面的距離,為求得更精確的干舷值,本文首先提出了基于Hough變換和均值漂移的船舶干舷值測量算法。此算法首先利用分析水面模型的方法來分離水面數(shù)據(jù)和船舶數(shù)據(jù),然后分別計(jì)算水面的高度和船舶甲板的高度,在計(jì)算船舶甲板高度的部分,用到了Hough變換和區(qū)域增長的方法提取出甲板數(shù)據(jù),并用均值漂移的方法計(jì)算船舶的甲板高度,最終獲得船舶的干舷值。在上述算法的基礎(chǔ)上,采用雙傳感器進(jìn)一步進(jìn)行數(shù)據(jù)結(jié)合,并優(yōu)化處理。本部分采用了基于Bayes分類的雙傳感器決策算法對數(shù)據(jù)進(jìn)行優(yōu)化處理,首先分析并提取出船舶的數(shù)據(jù)特征,然后再利用這些特征和Bayes算法來選擇哪一個(gè)傳感器測量的結(jié)果更接近于真實(shí)值。本文涉及的船舶超載、超吃水自動檢測系統(tǒng),已經(jīng)正式在長江流域岳陽分段運(yùn)行。該系統(tǒng)綜合了傳感器數(shù)據(jù)采集、視頻檢測、船舶干舷值測量三大主要模塊,實(shí)時(shí)將數(shù)據(jù)存儲后臺數(shù)據(jù)庫以便檢測。經(jīng)驗(yàn)證,該系統(tǒng)能夠?qū)崟r(shí)、可靠、穩(wěn)定的運(yùn)行。目前該系統(tǒng)每天平均檢測船舶120余艘,平均檢測誤差7 cm,滿足用戶對檢測系統(tǒng)的要求,且系統(tǒng)運(yùn)行穩(wěn)定可靠。這不僅有助于海事部門對船舶航行安全進(jìn)行管理,也有效遏制了船舶違法違規(guī)行為的發(fā)生。
[Abstract]:China is a large water transport country, inland waterways crisscross, the river reaches the sea, waterway traffic is very developed. In recent years, under the influence of domestic economic transformation and production adjustment, the phenomenon of inland river ship overloading has rebounded sharply, which has brought serious harm to the safety of water transportation. The ship freeboard value is regarded as the main criterion to measure whether the ship is overloaded or not, so the measurement and calculation of the ship freeboard value has become a hot research topic. In this paper, a ship freeboard measurement scheme based on laser scanning is proposed. The detailed research contents are as follows: the ship freeboard value is defined as the distance between the ship deck and the water surface, in order to obtain more accurate freeboard value, In this paper, a ship freeboard value measurement algorithm based on Hough transform and mean shift is proposed. The algorithm firstly uses the method of analyzing the water surface model to separate the surface data from the ship data, then calculates the height of the water surface and the height of the ship deck, respectively, in the part of calculating the height of the ship deck. The Hough transform and region growth are used to extract the deck data, and the mean shift method is used to calculate the ship's deck height. Finally, the freeboard value of the ship is obtained. On the basis of the above algorithm, two sensors are used to combine the data and optimize the processing. In this part, two sensor decision algorithms based on Bayes classification are used to optimize the data processing. Firstly, the data features of ships are analyzed and extracted. Then we use these features and Bayes algorithm to select which sensor measurement results are closer to the real value. The ship overload and overdraft automatic detection system involved in this paper has been officially operated in Yueyang section of the Yangtze River basin. The system integrates three main modules: sensor data acquisition, video detection, and ship freeboard value measurement. The data are stored in the background database in real time for detection. It is proved that the system can run in real time, reliable and stable. At present, the system detects more than 120 ships every day, and the average detection error is 7 cm, to meet the requirements of users, and the system runs stably and reliably. This not only helps the maritime department to manage the safety of ship navigation, but also effectively restrains the occurrence of illegal activities.
【學(xué)位授予單位】：哈爾濱工業(yè)大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：U693.4;TN249

【相似文獻(xiàn)】

相關(guān)會議論文 前1條

1 劉濤生;;對船舶超載現(xiàn)象的探討[A];江蘇省航海學(xué)會2004年優(yōu)秀論文集[C];2004年

相關(guān)碩士學(xué)位論文 前1條

1 趙英策;基于激光掃描的船舶干舷測量方法研究[D];哈爾濱工業(yè)大學(xué);2014年

，

本文編號：2329273