【摘要】：“遙感”這個(gè)詞是由E.L.Pruitt提出的，從20世紀(jì)60年代至今已經(jīng)發(fā)展成為一個(gè)非常先進(jìn)的檢測(cè)技術(shù)，在國(guó)民經(jīng)濟(jì)和國(guó)防領(lǐng)域得到了廣泛應(yīng)用，產(chǎn)生了巨大的社會(huì)和經(jīng)濟(jì)效益，成為了人類從多維和多角度認(rèn)識(shí)宇宙的新方法和新手段。在遙感圖像的采集過(guò)程中，總是不可避免的受到各種環(huán)境因素的影響，從而產(chǎn)生視覺效果差、分辨率低和亮度偏暗等不足，遙感傳感器的灰度范圍是采集到的遙感圖像的灰度所不能完全覆蓋的。因此，在提高圖像的對(duì)比度、突出一些局部細(xì)節(jié)等方面遙感圖像增強(qiáng)技術(shù)發(fā)揮著積極的作用。 圖像增強(qiáng)的方法大致可以分為兩種：空域圖像增強(qiáng)法和變換域圖像增強(qiáng)法�？沼驁D像增強(qiáng)顧名思義就是在圖像所在的空間域內(nèi)進(jìn)行增強(qiáng)處理，換句話說(shuō)就是直接對(duì)圖像中每個(gè)像素進(jìn)行某種處理。本文主要介紹一種類似于空域處理的方法--模糊算法，以及一些在傳統(tǒng)模糊算法基礎(chǔ)上的改進(jìn)算法。 變換域圖像增強(qiáng)法是將原始圖像變換到特定的變換域內(nèi)，修改處理變換域內(nèi)的相關(guān)系數(shù)來(lái)對(duì)圖像進(jìn)行相應(yīng)的增強(qiáng)。圖像從空域變換到變換域的方法有傅里葉變換、小波變換、Contourlet變換、NSCT變換、剪切波shearlet變換等。 本文將改進(jìn)的模糊算法分別應(yīng)用在空域圖像增強(qiáng)和變換域圖像增強(qiáng)中，提出了兩種遙感圖像的增強(qiáng)算法： 第一：采用改進(jìn)的模糊算法和一種空域圖像增強(qiáng)算法對(duì)遙感圖像進(jìn)行增強(qiáng)。由于傳統(tǒng)的Pal-King算法閾值難以確定，本文利用改進(jìn)的OTSU算法來(lái)自適應(yīng)獲取閾值，同時(shí)由于模糊增強(qiáng)是對(duì)全局進(jìn)行模糊處理的，細(xì)節(jié)信息增強(qiáng)效果不明顯，本章采用相對(duì)熵為判別標(biāo)準(zhǔn)，進(jìn)行領(lǐng)域信息的自適應(yīng)對(duì)比度增強(qiáng)。該算法在保持圖像原有亮度的基礎(chǔ)上同時(shí)增強(qiáng)了圖像細(xì)節(jié)。實(shí)驗(yàn)表明，，該算法能夠獲得很好的視覺效果和更加明顯的細(xì)節(jié)信息。 第二：首次將Shearlet變換引入到遙感圖像增強(qiáng)中提出了一種基于Shearlet變換的遙感圖像增強(qiáng)算法。對(duì)Shearlet變換產(chǎn)生的低頻系數(shù)和各尺度各方向的高頻系數(shù)分別進(jìn)行模糊對(duì)比增強(qiáng)和改進(jìn)的模糊增強(qiáng)，再經(jīng)過(guò)Shearlet反變換得到最終的增強(qiáng)結(jié)果。實(shí)驗(yàn)結(jié)果表明，該算法不僅在主觀上能獲得很好的視覺效果，同時(shí)在客觀評(píng)價(jià)中也擁有很好的指標(biāo)和性能。
[Abstract]:The term "remote sensing" was proposed by E.L.Pruitt and has been developed into a very advanced detection technology since the 1960s. It has been widely used in the field of national economy and national defense, and has produced enormous social and economic benefits. It has become a new method and means for human beings to understand the universe from multi-dimensional and multi-angle. In the process of remote sensing image acquisition, it is inevitable to be affected by various environmental factors, which result in poor visual effect, low resolution and dim brightness, etc. The range of gray scale of remote sensing sensor can not be completely covered by the grayscale of the collected remote sensing image. Therefore, remote sensing image enhancement technology plays an active role in improving image contrast and highlighting some local details. There are two methods for image enhancement: spatial image enhancement and transform domain image enhancement. Spatial image enhancement, as the name implies, is to enhance the image in the spatial domain, in other words, to directly process each pixel in the image. This paper mainly introduces a kind of similar spatial processing method-fuzzy algorithm, and some improved algorithms based on traditional fuzzy algorithm. Transform domain image enhancement method is to transform the original image into a specific transform domain and modify the correlation coefficients in the transform domain to enhance the image accordingly. The methods of image transform from spatial domain to transform domain include Fourier transform, wavelet transform, Contourlet transform, NSCT transform, shearlet transform of shear wave and so on. In this paper, the improved fuzzy algorithm is applied to spatial image enhancement and transform domain image enhancement, respectively. Two algorithms for remote sensing image enhancement are proposed. Firstly, an improved fuzzy algorithm and a spatial image enhancement algorithm are used to enhance the remote sensing image. Because the threshold of the traditional Pal-King algorithm is difficult to determine, this paper uses the improved OTSU algorithm to obtain the threshold from the adaptive algorithm. At the same time, because the fuzzy enhancement is a global fuzzy processing, the effect of detail information enhancement is not obvious. In this chapter, the relative entropy is used as the criterion to enhance the adaptive contrast of domain information. The algorithm enhances the image details on the basis of maintaining the original brightness of the image. Experiments show that the algorithm can obtain good visual effect and more obvious detail information. Second, Shearlet transform is introduced into remote sensing image enhancement for the first time. A remote sensing image enhancement algorithm based on Shearlet transform is proposed. The low frequency coefficients generated by Shearlet transform and the high frequency coefficients of each scale are enhanced by fuzzy contrast and improved fuzzy enhancement respectively. The final enhancement results are obtained by Shearlet inverse transformation. The experimental results show that the proposed algorithm can not only achieve a good visual effect subjectively, but also have a good performance in objective evaluation.
【學(xué)位授予單位】：新疆大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TP751

【參考文獻(xiàn)】

相關(guān)期刊論文 前8條

1 沙宇恒;劉芳;焦李成;;基于非下采樣Contourlet變換的SAR圖像增強(qiáng)[J];電子與信息學(xué)報(bào);2009年07期

2 柴勇;何友;曲長(zhǎng)文;;迭代離散Shearlet變換異類源遙感圖像融合[J];計(jì)算機(jī)工程與應(yīng)用;2011年03期

3 姜桃;趙春江;陳明;楊信廷;孫傳恒;;自適應(yīng)圖像模糊增強(qiáng)快速算法[J];計(jì)算機(jī)工程;2011年19期

4 王攀峰;杜云飛;周海芳;楊學(xué)軍;;基于復(fù)小波變換的遙感圖像并行融合算法[J];計(jì)算機(jī)工程與科學(xué);2008年03期

5 胡海智;孫輝;鄧承志;陳習(xí);柳枝華;占惠星;;基于Shearlet變換的圖像去噪算法[J];計(jì)算機(jī)應(yīng)用;2010年06期

6 趙嘉;孫輝;鄧承志;陳習(xí);;基于粒子群優(yōu)化的Shearlet自適應(yīng)圖像去噪[J];小型微型計(jì)算機(jī)系統(tǒng);2011年06期

7 胡海智;孫輝;鄧承志;陳習(xí);柳枝華;;全變差正則化的Shearlet收縮去噪[J];中國(guó)圖象圖形學(xué)報(bào);2011年02期

8 王明泉;馮曉夏;梁君婷;;基于直方圖均衡的射線圖像增強(qiáng)算法[J];中國(guó)科技論文在線;2011年01期

本文編號(hào)：2301962