【摘要】：近年來,隨著計算機視覺的發(fā)展和成熟,多視圖三維重建以其成本低廉,操作方便的優(yōu)勢倍受關(guān)注,目前已廣泛應(yīng)用于數(shù)字城市、醫(yī)療成像、虛擬現(xiàn)實等領(lǐng)域。圖像特征提取、特征匹配和光束法平差是多視圖三維重建的幾個關(guān)鍵步驟,但其中也在一些問題。首先,特征提取階段通常采用的SIFT算法并不能適用所有場景。其次,隨著GPU在計算能力和存儲器帶寬上展現(xiàn)出巨大的優(yōu)勢,基于GPU加速的多視圖三維重建算法成為研究的熱點。然而特征匹配算法遷移至GPU平臺的較少,并且由于硬件架構(gòu)的不同,并行算法的可靠性以及有限的顯存對光束法平差并行算法進行了極大的制約。為此本文展開了以下研究:(1)本文提出了一種Harris-Laplace特征結(jié)合SIFT描述的圖像配準(zhǔn)并行算法。特征提取階段,使用改進的Harris-Laplace算法提取出對圖像亮度、旋轉(zhuǎn)和尺度具有不變性的特征點,然后SIFT描述子對特征點進行描述。在特征點匹配階段則利用雙向匹配法和極線幾何約束對特征點進行粗匹配和精匹配。在分析該算法并行性的基礎(chǔ)上,采用CPU_GPU協(xié)同處理技術(shù),從任務(wù)的性質(zhì)和傳輸時間兩方面考慮,在CPU和GPU端進行合理的任務(wù)劃分,提升了整體算法的運行效率。(2)BA問題時常采用LM將其線性化,并用PCG算法求解法方程。本文先是利用PCG算法不需要顯式存儲法方程系數(shù)矩陣的特性,將其分解為簡單的雅各比矩陣和向量相乘的問題。然后在BA并行算法預(yù)處理過程中增加了一個過濾步驟,剔除了由數(shù)值轉(zhuǎn)換導(dǎo)致的錯誤點,保證精度的情況下最大限度發(fā)揮GPU高峰值的單精度浮點運算能力。最后在深入分析雅各比轉(zhuǎn)置矩陣與原矩陣聯(lián)系的基礎(chǔ)上,提出了一種光束法平差并行算法。對涉及到雅各比轉(zhuǎn)置矩陣的運算重新進行并行化設(shè)計,無需存儲雅各比轉(zhuǎn)置矩陣即可解決光束法平差問題。從實驗結(jié)果分析可知,本文提出Harris-Laplace特征結(jié)合SIFT描述的圖像配準(zhǔn)并行算法,提升了算法的整體效率,準(zhǔn)確度也得到可靠保證。本文提出光束法平差并行算法在獲得較好加速效果的同時,也大幅縮減了算法占用的顯存空間。借助GPU的強大的并行計算能力,多視圖三維重建實時性得到了很好的滿足。
[Abstract]:In recent years, with the development and maturity of computer vision, multi-view 3D reconstruction has been widely used in many fields such as digital city, medical imaging, virtual reality and so on because of its low cost and convenient operation. Image feature extraction, feature matching and beam adjustment are the key steps in multi-view 3D reconstruction, but there are also some problems. First of all, the SIFT algorithm usually used in the feature extraction phase is not suitable for all scenes. Secondly, with the huge advantages of GPU in computing power and memory bandwidth, the multi-view 3D reconstruction algorithm based on GPU has become a hot research topic. However, the migration of feature matching algorithm to GPU platform is less, and because of the different hardware architecture, the reliability of parallel algorithm and the limited display memory greatly restrict the parallel algorithm of beam adjustment. The main contents of this paper are as follows: (1) A parallel image registration algorithm based on Harris-Laplace feature and SIFT description is proposed in this paper. In the feature extraction phase, the improved Harris-Laplace algorithm is used to extract the feature points which are invariant to the brightness, rotation and scale of the image, and then the feature points are described by the SIFT descriptor. In the stage of feature point matching, two-way matching method and polar geometric constraint are used for coarse matching and fine matching of feature points. On the basis of analyzing the parallelism of the algorithm, the CPU_GPU cooperative processing technology is used to divide the tasks reasonably on the CPU and GPU sides, considering the nature of the task and the transmission time. The efficiency of the whole algorithm is improved. (2) LM is often used to linearize the BA problem and PCG algorithm is used to solve the equation. In this paper, we first decompose the PCG algorithm into a simple Yakubi matrix and a vector multiplication problem by taking advantage of the characteristic of the coefficient matrix of the equation without the need of explicit memory method. Then a filtering step is added to the pre-processing process of the BA parallel algorithm, which eliminates the error points caused by the numerical conversion and maximizes the single-precision floating-point operation capability of the GPU with high peak value under the condition of ensuring the accuracy. Finally, on the basis of deeply analyzing the relationship between Yakubi transposable matrix and original matrix, a parallel algorithm of beam adjustment is proposed. The parallel design of the operation involving Yakubi transposable matrix can solve the adjustment problem of beam method without storing Yakubi transposable matrix. From the analysis of the experimental results, this paper proposes a parallel image registration algorithm based on Harris-Laplace feature and SIFT description, which improves the overall efficiency of the algorithm and ensures the accuracy of the algorithm. In this paper, a parallel algorithm for beam adjustment is proposed, which not only achieves better acceleration effect, but also greatly reduces the memory space occupied by the algorithm. With the powerful parallel computing ability of GPU, the real-time performance of multi-view 3D reconstruction is satisfied.
【學(xué)位授予單位】：江西理工大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP391.41

【參考文獻】

相關(guān)期刊論文 前10條

1 桂生;劉洪;李飛;;簡化的混合域全波形反演方法及GPU加速[J];石油物探;2017年01期

2 陳潔;高志強;密保秀;陳會;;引入極線約束的SURF特征匹配算法[J];中國圖象圖形學(xué)報;2016年08期

3 安婷;賀一民;張志毅;;改進的雙向SIFT特征匹配算法[J];計算機工程與科學(xué);2016年01期

4 姜超;耿則勛;婁博;魏小峰;沈忱;;基于GPU的SIFT特征匹配算法并行處理研究[J];計算機科學(xué);2013年12期

5 劉鑫;許華榮;;基于GPU的特征點提取與匹配算法比較[J];計算機輔助設(shè)計與圖形學(xué)學(xué)報;2013年10期

6 聶俊嵐;陳歡歡;唐勇;張繼凱;;改進的GPU-SIFT特征提取與匹配算法[J];燕山大學(xué)學(xué)報;2013年02期

7 王蓓蕾;朱志良;孟t-;;基于CUDA加速的SIFT特征提取[J];東北大學(xué)學(xué)報(自然科學(xué)版);2013年02期

8 徐秋輝;佘江峰;宋曉群;肖鵬峰;;利用Harris-Laplace和SIFT描述子進行低空遙感影像匹配[J];武漢大學(xué)學(xué)報(信息科學(xué)版);2012年12期

9 張德好;劉青昆;宮利東;;GPU加速分子動力學(xué)模擬中的電荷分布計算[J];計算機應(yīng)用與軟件;2012年10期

10 肖漢;周清雷;張祖勛;;基于多GPU的Harris角點檢測并行算法[J];武漢大學(xué)學(xué)報(信息科學(xué)版);2012年07期

相關(guān)博士學(xué)位論文 前6條

1 肖漢;基于CPU+GPU的影像匹配高效能異構(gòu)并行計算研究[D];武漢大學(xué);2011年

2 田文;多視圖圖像的快速三維場景重建[D];華中科技大學(xué);2010年

3 朱清波;序列圖像三維重建方法研究[D];華中科技大學(xué);2010年

4 徐帆;無組織多視圖圖像的自動化三維場景重建[D];華中科技大學(xué);2007年

5 方磊;基于特征的圖像序列三維場景重建技術(shù)研究[D];華中科技大學(xué);2007年

6 陳付幸;基于非定標(biāo)圖像序列的三維重建關(guān)鍵技術(shù)研究[D];國防科學(xué)技術(shù)大學(xué);2005年

相關(guān)碩士學(xué)位論文 前3條

1 孟云;GPU加速的流體布料碰撞過程仿真[D];杭州電子科技大學(xué);2015年

2 陳鵬;基于GPU的圖像特征提取加速算法[D];復(fù)旦大學(xué);2013年

3 陳志雄;基于圖像配準(zhǔn)的SIFT算法研究與實現(xiàn)[D];武漢理工大學(xué);2008年

，

本文編號：2458898