本文關(guān)鍵詞： 智能交通系統(tǒng) 大氣能見度 暗通道先驗 交通監(jiān)控視頻　出處：《上海交通大學(xué)》2014年碩士論文　論文類型：學(xué)位論文

【摘要】：由于大氣能見度對交通中的人和車的行為有重要的影響,道路能見度的監(jiān)測一直是智能交通系統(tǒng)的一個重要組成部分。惡劣的能見度狀況不僅不利于人們的健康,而且容易引發(fā)交通事故,威脅人們的出行安全,因此對于低能見度的檢測與預(yù)警尤為重要。傳統(tǒng)的能見度測量方法主要包括人工目測方法和儀器測量方法。前者主觀性強,精度差；后者操作復(fù)雜,成本高昂。因此,改進能見度測量技術(shù)具有重要的現(xiàn)實意義與應(yīng)用價值。 本文在對能見度相關(guān)的物理模型研究的基礎(chǔ)上,討論了不同應(yīng)用場景下的基于圖像理解的能見度估計方法,論述了其相關(guān)的測量原理,并提出了一個基于交通視頻的能見度等級估計方法。具體的說,該方法首先從視頻圖像中提取出全局平均梯度和暗通道比率特征,并由此建立能見度等級的估計模型,最后用一個直方圖校正模塊和一個晴天檢測模塊對估計結(jié)果進行修正。本文通過大量的實驗來驗證了該方法的有效性。 此外,本文還自主研發(fā)了一個道路環(huán)境信息的集成平臺,以更全面地反映道路環(huán)境信息。其集成的信息包括能見度等級、溫度以及濕度。該平臺主要包含3個方面的功能：第一,根據(jù)輸入的道路監(jiān)控視頻估計大氣能見度等級；第二,通過USB接口讀取溫濕度傳感器采集的溫度和濕度信息；第三,將獲取的溫度、濕度以及能見度等信息通過視頻水印的方式進行集成與可視化。
[Abstract]:Atmospheric visibility has an important impact on the behavior of people and cars in traffic. Visibility monitoring is an important part of intelligent transportation system. Poor visibility is not only harmful to people's health, but also easy to cause traffic accidents and threaten people's travel safety. Therefore, it is very important for low visibility detection and early warning. The traditional visibility measurement methods mainly include manual visual measurement method and instrument measurement method. The former has strong subjectivity and poor precision. The latter is complex in operation and high in cost. Therefore, improving visibility measurement technology has important practical significance and application value. Based on the study of the physical model of visibility correlation, this paper discusses the visibility estimation methods based on image understanding in different application scenarios, and discusses the related measurement principles. A visibility level estimation method based on traffic video is proposed. In detail, the global average gradient and dark channel ratio feature are extracted from the video images. Finally, a histogram correction module and a sunny detection module are used to correct the estimation results. A large number of experiments are carried out to verify the effectiveness of the proposed method. In addition, this paper also developed a road environment information integration platform to reflect the road environment information more comprehensively. The integrated information includes visibility level. Temperature and humidity. The platform mainly includes three functions: firstly, the visibility level of atmosphere is estimated according to the input video of road surveillance; Secondly, the temperature and humidity information collected by the temperature and humidity sensor is read through the USB interface. Thirdly, the information of temperature, humidity and visibility is integrated and visualized by video watermarking.
【學(xué)位授予單位】：上海交通大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：U495;TP391.41

【參考文獻】

相關(guān)期刊論文 前10條

1 王國鋒;宋鵬飛;張?zhí)N靈;;智能交通系統(tǒng)發(fā)展與展望[J];公路;2012年05期

2 李學(xué)彬;遲如利;徐青山;魏合理;;利用PM10和PM2．5反演大氣能見度[J];光學(xué)技術(shù);2008年04期

3 饒瑞中;;大氣中的視覺和大氣能見度[J];光學(xué)學(xué)報;2010年09期

4 陳曉婷;魯昌華;張玉均;劉文清;戴龐達;常峰;朱超;;數(shù)字?jǐn)z像能見度儀的白天能見度算法設(shè)計[J];電子測量與儀器學(xué)報;2013年04期

5 邢向楠;崔巖梅;張富根;謝邦力;;能見度測量技術(shù)現(xiàn)狀及發(fā)展趨勢綜述[J];計測技術(shù);2010年05期

6 李勃;董蓉;陳啟美;;無需人工標(biāo)記的視頻對比度道路能見度檢測[J];計算機輔助設(shè)計與圖形學(xué)學(xué)報;2009年11期

7 張江,趙黎,楊士強;視頻水印技術(shù)綜述[J];計算機應(yīng)用;2005年04期

8 謝興生,陶善昌,周秀驥;數(shù)字?jǐn)z像法測量氣象能見度[J];科學(xué)通報;1999年01期

9 胡江華，，李漢斌，周建勛，張保民;夜間氣象能見度的測量研究[J];南京氣象學(xué)院學(xué)報;1995年04期

10 許茜;殷緒成;李巖;郝紅衛(wèi);曹曉鐘;;基于圖像理解的能見度測量方法[J];模式識別與人工智能;2013年06期

本文編號：1456423