即時的SIFT特徵點擷取之低記憶體硬體設計

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：9

、訪客IP：3.145.206.169

姓名

王瑞智(Rui-Zhi Wang) 查詢紙本館藏

畢業系所

電機工程學系

論文名稱

即時的SIFT特徵點擷取之低記憶體硬體設計
(Hardware architecture for real-time SIFT extraction with reduced memory)

相關論文

★ 即時的人臉偵測與人臉辨識之門禁系統	★ 具即時自動跟隨功能之自走車
★ 應用於多導程心電訊號之無損壓縮演算法與實現	★ 離線自定義語音語者喚醒詞系統與嵌入式開發實現
★ 晶圓圖缺陷分類與嵌入式系統實現	★ 語音密集連接卷積網路應用於小尺寸關鍵詞偵測
★ G2LGAN: 對不平衡資料集進行資料擴增應用於晶圓圖缺陷分類	★ 補償無乘法數位濾波器有限精準度之演算法設計技巧
★ 可規劃式維特比解碼器之設計與實現	★ 以擴展基本角度CORDIC為基礎之低成本向量旋轉器矽智產設計
★ JPEG2000靜態影像編碼系統之分析與架構設計	★ 適用於通訊系統之低功率渦輪碼解碼器
★ 應用於多媒體通訊之平台式設計	★ 適用MPEG 編碼器之數位浮水印系統設計與實現
★ 適用於視訊錯誤隱藏之演算法開發及其資料重複使用考量	★ 一個低功率的MPEG Layer III 解碼器架構設計

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 ( 永不開放)

摘要(中)

在電腦視覺領域中物件匹配及物件辨識一直是廣為研究的題目，其方法主要可透過建立影像中的特徵點再進行特徵點的比對，在眾多的特徵點擷取演算法中，SIFT(Scale invariant feature transform)被認為是效果最好的演算法之一，對於影像縮放、旋轉、變形及光影變化皆能保持不變性，然而其高效及穩定的背後需要付出的是複雜的運算及大量的記憶體空間，這使得SIFT演算法難以達到即時運算（Real-time）。
為了達到即時運算及降低記憶體需求，我們在本文中提出了一個新的SIFT硬體架構設計，並對SIFT演算法進行改良。在建構高斯金字塔時我們捨棄原先cascade運算取代為平行運算，雖然增加運算量但可降低暫存記憶體並增加速度；而對於第一組(octave)影像不進行放大取樣，採用較小的尺度空間來模擬高頻的空間頻率；特徵點定位則是選擇改用適當閥值濾除過小的極值，此外整體演算法是基於區塊影像進行運算，透過以上優化能夠大幅降低設計中所需的內部記憶體。我們將硬體架構透過FPGA (Xilinx Artix7)進行驗證，並採用TSMC90nm製程進行實做。實驗結果顯示，在解析度為1280x720的影像中，工作頻率可達152MHz，處理速度可達35.6 frames/s，內部記憶體需求為237Kb。

摘要(英)

Object matching and identification are the popular research in the field of computer vision. The methods are mainly based on feature matching. In many feature extraction algorithms, SIFT (Scale invariant feature transform) is considered to be one of the best algorithms. It maintains the robustness for image scaling, rotation, deformation, and light changes. However, it require complex calculations and large amounts of memory that makes it difficult to achieve real-time.
In order to achieve real-time operations and reduce memory requirements, we propose a new SIFT hardware architecture and improve the SIFT algorithm in this thesis. In the step of constructing Gaussian pyramids, the original cascade operations is substituted by parallel operations. Although it increase the amount of calculations, we can reduce the amount of temporary memory and increase the speed. For the first octave images, we adapt a smaller scale space to simulate the high spatial frequency instead of upsampling image. The step of keypoint localization is replaced by appropriate threshold to filter out the bad extremum. In addition, the overall algorithm is based on segment which can be greatly reduced the internal memory in the design. We verified the hardware architecture on FPGA (Xilinx Artix7) and implemented by TSMC90nm process. Experimental results show that for an image with a resolution of 1280x720, the operating frequency can reach 152 MHz, the processing speed can reach 35.6 frames/s, and the internal memory requirement is 237 Mbits.

關鍵字(中)

★ 尺度不變特徵轉換

關鍵字(英)

★ SIFT

論文目次

致謝 i
摘要 v
Abstract vi
Table of contents vii
List of Figures ix
List of Tables xi
Chapter I Introduction 1
1.1 Introduction 1
1.2 Motivation 5
1.3 Thesis Organization 6
Chapter II Related work 7
2.1 PCA-SIFT 7
2.2 SURF 8
2.3 GLOH 10
2.4 Summary 11
Chapter III SIFT 12
3.1 Detection of scale-space extrema 13
3.2 Accuracy keypoint localization 16
3.3 Orientation assignment 19
3.4 Keypoint descriptor 21
Chapter IV Hardware architecture 23
4.1 Overall hardware architecture 24
4.2 The Architecture of Scale-space Construction 26
4.2.1 Reduce initial scale of Gaussian pyramid 26
4.2.2 Parallel Gaussian filter 27
4.3 The Architecture Accuracy keypoint localization 31
4.3.1 Eliminate low contrast keypoint 32
4.4 The Architecture Descriptor Assignment 33
4.4.1 CORDIC algorithm 34
Chapter V Experiment Result 36
5.1 Experiment Environment setting 36
5.2 Invariant Test Result 37
5.3 Hardware Architecture Implementation 40
Chapter VI Conclusion 43
Reference 44

參考文獻

[1] J.Canny, “A computational approach to edge detection.,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 8, no. 6, pp. 679–698, 1986.
[2] C.Harris andM.Stephens, “A Combined Corner and Edge Detector,” Procedings Alvey Vis. Conf. 1988, p. 23.1-23.6, 1988.
[3] S. M.Smith andJ. M.Brady, “SUSAN---A New Approach to Low Level Image Processing,” Int. J. Comput. Vis., vol. 23, no. 1, pp. 45–78, 1997.
[4] E.Rosten andT.Drummond, “Machine learning for high-speed corner detection,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2006, vol. 3951 LNCS, pp. 430–443.
[5] D. G.Lowe, “Distinctive image features from scale invariant keypoints,” Int. J. Comput. Vis., vol. 60, no. 2, pp. 91–11020042, 2004.
[6] H.Bay, T.Tuytelaars, andL.VanGool, “SURF: Speeded up robust features,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2006, vol. 3951 LNCS, pp. 404–417.
[7] A. K.Jain, J. E.Lee, R.Jin, andN.Gregg, “Content-based image retrieval: An application to tattoo images,” in Proceedings - International Conference on Image Processing, ICIP, 2009, pp. 2745–2748.
[8] A.Kuffner andA.Robles-Kelly, “Image feature evaluation for contents-based image retrieval,” Proc. HCSNet Work. …, vol. 56, no. VisHCI, pp. 2–6, 2006.
[9] S.-W.Ha andY.-H.Moon, “Multiple Object Tracking Using SIFT Features and Location Matching,” Int. J. Smart Home, vol. 5, no. 4, 2011.
[10] H.Zhou, Y.Yuan, andC.Shi, “Object tracking using SIFT features and mean shift,” Comput. Vis. Image Underst., vol. 113, no. 3, pp. 345–352, Mar.2009.
[11] P.Azad, T.Asfour, andR.Dillmann, “Combining Harris interest points and the SIFT descriptor for fast scale-invariant object recognition,” in 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2009, 2009, pp. 4275–4280.
[12] G.Flitton, T.Breckon, andN.Megherbi Bouallagu, “Object Recognition using 3D SIFT in Complex CT Volumes,” in Procedings of the British Machine Vision Conference 2010, 2010, p. 11.1-11.12.
[13] Liu Jinxia andQiu Yuehong, “Application of SIFT feature extraction algorithm on the image registration,” in IEEE 2011 10th International Conference on Electronic Measurement & Instruments, 2011, pp. 177–180.
[14] Yan Ke andR.Sukthankar, “PCA-SIFT: a more distinctive representation for local image descriptors,” in Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004., 2004, vol. 2, pp. 506–513.
[15] K.Mikolajczyk, K.Mikolajczyk, C.Schmid, andC.Schmid, “A performance evaluation of local descriptors,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 27, no. 10, pp. 1615–1630, 2005.
[16] M. (Eds. .Awad, Ali Ismail, Hassaballah, Image Feature Detectors and Descriptors: Foundations and Applications. .
[17] T.Lindeberg, “Scale-space theory: A basic tool for analysing structures at diierent s c a l e s,” J. Appl. Stat., vol. 21, no. 2, pp. 225–270, 1994.
[18] K.Mikolajczyk andC.Schmid, “Scale & affine invariant interest point detectors,” Int. J. Comput. Vis., vol. 60, no. 1, pp. 63–86, 2004.
[19] V.Bonato, E.Marques, andG.Constantinides, “A Parallel Hardware Architecture for Scale and Rotation Invariant Feature Detection,” IEEE Trans. Circuits Syst. Video Technol., vol. 18, no. 12, pp. 1703–1712, 2008.
[20] “CORDIC.” [Online]. Available: https://en.wikibooks.org/wiki/Digital_Circuits/CORDIC.
[21] F. C.Huang, S. Y.Huang, J. W.Ker, andY. C.Chen, “High-performance SIFT hardware accelerator for real-time image feature extraction,” IEEE Trans. Circuits Syst. Video Technol., vol. 22, no. 3, pp. 340–351, 2012.
[22] L.Chiu andT.Chang, “Fast sift design for real-time visual feature extraction,” Image Process. IEEE …, vol. 22, no. 8, pp. 3158–67, 2013.
[23] J.Yum, C. H.Lee, J. S.Kim, andH. J.Lee, “A novel hardware architecture with reduced internal memory for real-time extraction of SIFT in an HD video,” IEEE Trans. Circuits Syst. Video Technol., vol. 26, no. 10, pp. 1943–1954, 2016.

指導教授

蔡宗漢(Tsung-Han Tsai)

審核日期

2018-5-8

推文