本文選題：白光效應(yīng) + 偏振圖像　； 參考：《長春理工大學(xué)》2016年碩士論文

【摘要】：當(dāng)用激光主動成像設(shè)備在夜間采集圖像時,有時會出現(xiàn)在獲取的圖像上有明顯的白色光斑的情況,我們稱這種現(xiàn)象為白光效應(yīng)。白光效應(yīng)的發(fā)生會使場景中的部分信息被掩蓋,致使我們獲取的信息不完整,這就對圖像的采集和后續(xù)的圖像處理工作造成了極大的干擾,所以對白光效應(yīng)這樣一種常見的現(xiàn)象展開研究是很有必要的。本文主要是以能夠有效消除白光效應(yīng)的影響,獲取包含完整場景信息的可見光圖像為目的而進(jìn)行的相關(guān)方面的研宄。文中基于激光的偏振特性,充分利用偏振圖像和光強(qiáng)圖像的互補(bǔ)性,運(yùn)用圖像融合技術(shù),針對研究內(nèi)容和目的依次提出了三種圖像處理方法:基于自適應(yīng)加權(quán)與IHS相結(jié)合的研究方法、基于DWT的多分辨率分析研究方法、基于圖像分割與光照補(bǔ)償技術(shù)的研究方法。論文中對每種方法的理論基礎(chǔ)、實(shí)現(xiàn)步驟、仿真結(jié)果都進(jìn)行詳細(xì)的研究與闡述,并通過與其它相關(guān)算法進(jìn)行對比分析,證明了本論文提出的算法的有效性和優(yōu)越性。文中對每種研究方法及其他文獻(xiàn)中的相關(guān)算法都做了仿真,仿真結(jié)果表明,雖然本文的幾種處理方法的原理不同,處理后的結(jié)果也存在差異,但是融合后的圖像都成功消除了白光效應(yīng)的影響,比其它文獻(xiàn)中的仿真結(jié)果要好。本文的研究方法將非偏圖像和偏振圖像的場景信息成功的綜合在了一起,取長補(bǔ)短,得到了更加清晰,含有的信息更多的圖像,切實(shí)達(dá)到了研究的要求和目的。
[Abstract]:When laser active imaging equipment is used to collect images at night, there are obvious white spots on the acquired images, which is called the white light effect. The occurrence of the white light effect will cover up part of the information in the scene, which will lead to incomplete information, which will cause great interference to the image acquisition and subsequent image processing. So it is necessary to study the white light effect as a common phenomenon. The purpose of this paper is to effectively eliminate the white light effect and obtain the visible light image which contains the complete scene information. Based on the polarization characteristics of laser, this paper makes full use of the complementarity between polarized image and intensity image, and applies image fusion technology. According to the content and purpose of the research, three image processing methods are put forward: one is based on adaptive weighting and the other is IHS, the other is multi-resolution analysis based on DWT, and the research method based on image segmentation and illumination compensation. In this paper, the theoretical basis, implementation steps, simulation results of each method are studied and elaborated in detail, and compared with other related algorithms, the effectiveness and superiority of the algorithm proposed in this paper are proved. The simulation results show that although the principles of several methods in this paper are different, the results are different. However, the fusion images successfully eliminate the white light effect, which is better than the simulation results in other literatures. The research method of this paper successfully synthesizes the scene information of the unbiased image and the polarized image, which can make up for each other, get more clear and contain more information, and achieve the requirement and purpose of the research.
【學(xué)位授予單位】：長春理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2016
【分類號】：TP391.41

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 孟勇,洪丹輝,毛丹;測度熵在圖像紋理分析中的應(yīng)用[J];計算機(jī)應(yīng)用與軟件;2000年08期

2 吳濤;秦昆;;圖像紋理特征數(shù)據(jù)挖掘的理論與方法探討[J];計算機(jī)時代;2006年08期

3 方玲玲;王相海;;圖像挖掘研究[J];計算機(jī)科學(xué);2009年08期

4 高振宇;楊曉梅;龔劍明;金海;;圖像復(fù)雜度描述方法研究[J];中國圖象圖形學(xué)報;2010年01期

5 劉勇,施萬昌,徐玉蘭;圖像差異的分析與識別[J];復(fù)旦學(xué)報(自然科學(xué)版);2000年05期

6 羅l，

本文編號：1950118