簡易檢索 / 詳目顯示
| 研究生： |
辜致翔 |
|---|---|
| 論文名稱： |
基於三元化特徵描述子之行動影像識別機制 Image recognition on mobile devices using ternary feature descriptors |
| 指導教授： | 廖文宏 |
| 學位類別： |
碩士
Master |
| 系所名稱： |
理學院 - 資訊科學系 |
| 論文出版年： | 2014 |
| 畢業學年度： | 102 |
| 語文別： | 中文 |
| 論文頁數： | 65 |
| 中文關鍵詞： | 三元化特徵描述子 、二元化特徵描述子 、行動影像辨識 |
| 外文關鍵詞： | ternary feature descriptors, binary feature descriptor, mobile image recognition, feature recognition |
| 相關次數： | 點閱：109 下載：4 |
| 分享至: |
| 查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
現在因應科技的發展，行動裝置計算能力和應用也成長快速，現行非常多應用需要結合各種不同的感應器或相機等來配合使用，然而這些裝置在不同的狀況下會有一些限制，在無法改變外在的環境下，勢必需要靠軟體去做補強或修正。
本論文針對在有限的計算和存儲資源的移動裝置平台上進行物體的偵測與追踪。為了達到此一目標，我們提出了一個型態的圖像特徵，稱為區域三元描述子（Local Ternary Descriptors），期望在時間複雜度、抗噪性和準確率各個面向取得一個較佳的平衡。LTD是基於區域二元描述子（Local Binary Descriptors）所衍生出來的方法，如BRIEF，BRISK，FREAK。而使用三進位制編碼方法的動機在於，三元化處理可以減輕因在LBD的簡單threshold處理過後所產生的一些問題。而類似於LBD地方在於，LTDS之間的距離可以很容易地使用Hamming distance計算。實驗數據及比較分析後證明，本論文提出的區域三元描述子可以在雜訊環境的條件下表現出優異的效果。
The rapid advances of information and communication technology have brought about the prevalence of mobile devices. Diverse applications on smartphones have emerged accordingly. Interactive media and augmented reality are two well-known examples that utilize these devices as an interface to present digital content to the users. Effective interface design is therefore a critical factor to guarantee satisfactory user experience.
In this thesis, we address the detection and tracking of objects on mobile platforms with limited computation and storage resources. To strike a good balance among feature complexity, noise immunity and detection rate, we propose a novel class of image feature known as local ternary descriptors (LTD). LTDs are extensions of local binary descriptors (LBD) such as BRIEF, BRISK, and FREAK. The motivation for using ternary representation lies in the observation that the ternarization process can alleviate some problems caused by simple thresholding in LBD. Similar to LBD, the distance between two LTDs can be easily computed using Hamming distance. Experimental results and comparative analysis indicate that the proposed descriptor can achieve superior performance under noisy conditions.
1. 第一章 緒論 1
1.1 研究背景 1
1.2 目的 2
1.3 流程架構與方法 2
2. 第二章 相關研究 5
2.1. 特徵擷取 5
2.1.1. SUSAN(Smallest Univalue Segment Assimilating Nucleus) 6
2.1.2. FAST (features from accelerated segment test) 7
2.2. 區域興趣點偵測 (regions of interest or interest points) 8
2.2.1. 高斯拉普拉斯轉換 (The Laplacian of Gaussian, LoG) 8
2.2.2. 高斯差 (Difference of Gaussians, DoG) 10
2.3. 特徵描述子 10
2.3.1. 尺度不變特徵轉換 (Scale-invariant feature transform, SIFT) 11
2.3.2. PCA-SIFT (Principle Component Analysis-SIFT) 13
2.3.3. GLOH (Gradient Location-Orientation Histogram) 13
2.3.4. SURF (Speeded Up Robust Features) 14
2.3.5. BRIEF (Binary Robust Independent Elementary Features) 16
2.3.6. ORB (Oriented FAST and Rotated BRIEF) 17
2.3.7. BRISK (Binary Robust Invariant Scalable Keypoints) 18
2.3.8. Fast Retina Keypoint (FREAK) 19
3. 第三章 區域三元化描述子 23
3.1. 二元化特徵值討論歸納 23
3.2. 二元化特徵值實驗分析 26
3.3. 區域三元化圖型 27
4. 第四章 實驗結果與數據 31
4.1. 影像雜訊 31
4.2. 抗噪性測試 36
4.3. 演算法三元化特徵值套用實驗 43
4.4. 實驗統整分析 55
5. 第五章 擴增實境應用 56
6. 第六章 結論與未來研究方向 58
參考文獻 59
A. 附錄 61
[1] Canny, J. "A Computational Approach To Edge Detection". IEEE Trans. Pattern Analysis and Machine Intelligence 8 (6): 679–714, 1986.
[2] Scharr, Hanno. Dissertation (in German), Optimal Operators in Digital Image Processing, 2000.
[3] C. Harris and M. Stephens. "A combined corner and edge detector". Proceedings of the 4th Alvey Vision Conference. pp. pages 147—151, 1988.hi
[4] S. M. Smith and J. M. Brady . "SUSAN - a new approach to low level image processing". International Journal of Computer Vision 23 (1): 45–78, 1997.
[5] M. Trajkovic and M. Hedley. "Fast corner detection". Image and Vision Computing 16 (2): 75–87, 1998.
[6] D. Lowe. "Distinctive Image Features from Scale-Invariant Keypoints", International Journal of Computer Vision 60 (2), 2004.
[7] DG Lowe, Object recognition from local scale-invariant features. Computer Vision, 1999. The Proceedings of the Seventh IEEE International Conference, vol 2, 1150-1157, 1999.
[8] H Bay, T Tuytelaars, L Van Gool. Surf: Speeded up robust features. Computer Vision–ECCV 2006.
[9] Calonder, Michael, et al. "BRIEF: binary robust independent elementary features." Computer Vision–ECCV 2010. Springer Berlin Heidelberg, 2010. 778-792.
[10] Rublee, E., Rabaud, V., Konolige, K., & Bradski, G. (2011, November). ORB: an efficient alternative to SIFT or SURF. In Computer Vision (ICCV), 2011 IEEE International Conference on (pp. 2564-2571). IEEE.
[11] Leutenegger, Stefan, Margarita Chli, and Roland Y. Siegwart. "BRISK: Binary robust invariant scalable keypoints." Computer Vision (ICCV), 2011 IEEE International Conference on. IEEE, 2011.
[12] A Alahi, R Ortiz, P Vandergheynst. FREAK: Fast Retina Keypoint. Computer Vision and Pattern Recognition (CVPR), 510 – 517, 2012
[13] J Heinly, E Dunn, JM Frahm Comparative Evaluation of Binary Features Computer Vision–ECCV 2012
[14] Cha, S. H., Yoon, S., & Tappert, C. C. (2005). Enhancing binary feature vector similarity measures.
[15] Miksik, O., & Mikolajczyk, K. (2012, November). Evaluation of local detectors and descriptors for fast feature matching. In Pattern Recognition (ICPR), 2012 21st International Conference on (pp. 2681-2684). IEEE.
[16] Mair, E., Hager, G. D., Burschka, D., Suppa, M., & Hirzinger, G. (2010). Adaptive and generic corner detection based on the accelerated segment test. In Computer Vision–ECCV 2010 (pp. 183-196). Springer Berlin Heidelberg.
[17] K. Mikolajczyk and C. Schmid. A performance evaluation of local descriptors. IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI), 2:1115–1125, 2005. 2, 5
[18] 楊挺榮(2010) 基於延展式區域三元化徒刑之特徵描述子
