【摘要】：植株的三維模型重建以及植株的三維模型可視化是植物形態(tài)學(xué)領(lǐng)域的研究熱點,而植株的三維點云數(shù)據(jù)是植株三維重建的基礎(chǔ),因此如何方便高效地獲取三維點云成為了重點研究內(nèi)容。目前大多數(shù)的點云獲取系統(tǒng)都是基于計算機實現(xiàn)的,此類系統(tǒng)因安裝在固定設(shè)備上,不能滿足移動用戶的需求。隨著移動終端硬件水平的提升和無線通信網(wǎng)絡(luò)的不斷成熟,基于Android移動端的三維點云獲取系統(tǒng)成為一種可行的解決方案。此方案能夠滿足人們更快捷、更方便地進(jìn)行點云提取工作的需求�；诖�,本文提出了一套基于移動端的植株三維點云獲取方案,并依據(jù)該方案構(gòu)建出了一套完整的基于移動端的植株三維點云獲取系統(tǒng),最后對系統(tǒng)進(jìn)行了測試。本文主要研究內(nèi)容如下:(1)提出了一種基于移動端的植株三維點云獲取方案。針對PC端獲取點云靈活性差和移動端處理速度慢、內(nèi)存性能局限性強等問題,提出了采用客戶端-服務(wù)器系統(tǒng)架構(gòu)開發(fā)基于移動端的植株三維點云獲取系統(tǒng)的方案。此方案旨在通過移動端獲取物體的圖像序列,并上傳至服務(wù)器進(jìn)行點云計算,最后服務(wù)器傳回點云至移動端進(jìn)行顯示。(2)開發(fā)了基于移動端的植株三維點云獲取系統(tǒng)。針對傳統(tǒng)點云獲取技術(shù)成本高、靈活性差等問題,采用基于移動端的植株三維點云獲取方案開發(fā)了本系統(tǒng)。首先,從技術(shù)、理論、經(jīng)濟三方面對系統(tǒng)進(jìn)行可行性分析。接著,通過對功能需求和性能需求的分析,進(jìn)一步確定了系統(tǒng)要完成的任務(wù)。然后,通過需求分析對系統(tǒng)進(jìn)行了總體設(shè)計和模塊設(shè)計。最后,基于Android開發(fā)技術(shù)、服務(wù)器搭建技術(shù)、SIFT算法、Bundler算法和PMVS算法對系統(tǒng)的各個功能模塊進(jìn)行實現(xiàn)。最終開發(fā)了一套完整的基于移動端的植株三維點云獲取系統(tǒng)。最后,為了驗證本系統(tǒng)是否滿足功能需求和性能需求,對系統(tǒng)進(jìn)行測試。采用植株和非植株兩類測試用例對系統(tǒng)進(jìn)行了功能測試,通過控制不同變化因素對系統(tǒng)進(jìn)行了處理時間和三維點數(shù)量兩方面的性能測試。測試結(jié)果顯示,本系統(tǒng)對于不同特征的植株和非植株物體均能較好地獲取其點云數(shù)據(jù),且當(dāng)圖像數(shù)量為30、圖像分辨率為1600×1200時,系統(tǒng)處理時間保持在20分鐘之內(nèi),這滿足預(yù)期的功能需求和性能需求。
[Abstract]:The 3D model reconstruction and the visualization of plant 3D model are the research focus in the field of plant morphology, and the 3D point cloud data of plant is the basis of plant 3D reconstruction. Therefore, how to obtain three-dimensional point cloud conveniently and efficiently has become the focus of research. At present, most of the point cloud acquisition systems are based on computer, such systems can not meet the needs of mobile users because they are installed on fixed devices. With the improvement of hardware level of mobile terminal and the maturity of wireless communication network, 3D point cloud acquisition system based on Android mobile terminal becomes a feasible solution. This scheme can meet the needs of point cloud extraction more quickly and conveniently. Based on this, this paper proposes a plant 3D point cloud acquisition scheme based on mobile end, and constructs a complete plant 3D point cloud acquisition system based on mobile end. Finally, the system is tested. The main contents of this paper are as follows: (1) A scheme of plant 3D point cloud acquisition based on mobile end is proposed. Aiming at the problems of poor flexibility of point cloud acquisition on PC side, slow processing speed on mobile side and strong limitation of memory performance, a scheme of developing plant 3D point cloud acquisition system based on mobile end is proposed by using client-server system architecture. The purpose of this scheme is to obtain the image sequence of the object through the mobile end and upload it to the server for point cloud computing. Finally, the server sends the point cloud back to the mobile side for display. (2) A plant 3D point cloud acquisition system based on the mobile end is developed. Aiming at the problems of high cost and poor flexibility of traditional point cloud acquisition technology, this system is developed by using the scheme of plant 3D point cloud acquisition based on mobile end. Firstly, the feasibility of the system is analyzed from three aspects: technology, theory and economy. Then, through the analysis of the functional and performance requirements, the tasks to be completed by the system are further determined. Then, through the requirement analysis, the overall design and module design of the system are carried out. Finally, based on the Android development technology, server construction technology, SIFT algorithm, Bundler algorithm and PMVS algorithm to implement each functional module of the system. Finally, a complete plant 3D point cloud acquisition system based on mobile end is developed. Finally, in order to verify whether the system meets the functional requirements and performance requirements, the system is tested. Plant and non-plant test cases were used to test the function of the system. The processing time and the number of 3D points were tested by controlling different factors. The test results show that the system can obtain the point cloud data for different plant and non-plant objects, and when the number of images is 30 and the resolution of image is 1600 脳 1200, the processing time of the system is kept within 20 minutes. This meets the expected functional and performance requirements.
【學(xué)位授予單位】：西北農(nóng)林科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：Q94;TP391.41

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 馬福峰;耿楠;張志毅;;基于鄰域幾何特征約束的植株三維形態(tài)配準(zhǔn)方法研究[J];計算機應(yīng)用與軟件;2016年09期

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

3 韓笑;廖粵峰;;基于平面二次曲線的非單視點全景相機標(biāo)定方法[J];計算機與現(xiàn)代化;2015年07期

4 李博陽;耿楠;張志毅;;基于機器視覺的激光條紋掃描系統(tǒng)[J];計算機仿真;2015年06期

5 鄧衛(wèi)成;;信息系統(tǒng)服務(wù)器資源需求估算方法研究[J];通訊世界;2015年02期

6 徐偉亮;龍文麗;陳輝;陳雄;;基于人因工程的智能手機操作系統(tǒng)用戶體驗評估[J];中國科技信息;2015年01期

7 裴驕陽;張志毅;袁聰聰;;基于單目視覺幾何構(gòu)建的三維掃描系統(tǒng)[J];計算機與現(xiàn)代化;2014年06期

8 劉剛;司永勝;馮娟;;農(nóng)林作物三維重建方法研究進(jìn)展[J];農(nóng)業(yè)機械學(xué)報;2014年06期

9 宋妍;茍永剛;鄭坤;;一種新的遙感影像角點提取方法[J];遙感信息;2014年02期

10 郭振成;梁鳳梅;;一種改進(jìn)的SIFT特征點匹配算法[J];電視技術(shù);2014年01期

相關(guān)會議論文 前1條

1 張飛;;相機標(biāo)定方法綜述[A];2011年全國通信安全學(xué)術(shù)會議論文集[C];2011年

相關(guān)碩士學(xué)位論文 前4條

1 羅其朝;基于Android的圖像識別客戶端系統(tǒng)研究[D];海南大學(xué);2014年

2 范琳;基于圖像處理的點云獲取及處理技術(shù)研究[D];大連工業(yè)大學(xué);2014年

3 劉華洲;基于智能手機傳感器的圖像匹配技術(shù)研究[D];西安電子科技大學(xué);2013年

4 周春江;基于L系統(tǒng)的虛擬植物生長的模擬研究[D];重慶大學(xué);2005年

，

本文編號：2370452