吳家麟臺灣大學：資訊工程學研究所陳駿丞Chen, Jun-ChengJun-ChengChen2007-11-262018-07-052007-11-262018-07-052006http://ntur.lib.ntu.edu.tw//handle/246246/53978這篇論文提出一種新的瀏覽媒介，複合式拼貼幻燈秀，以拼貼的方式展示照片，並配合襯底音樂的節拍播放。與傳統的照片幻燈秀比較，多張具相似特性的照片妥善地安排於相同版面中並播放。由科技寫作激發的靈感，每個版面配置由一張主題照片與多張修飾照片所組成。基於這想法，本幻燈秀系統由三個主要部分構成︰影像群集，音樂分析，版面配置。受限於展示的空間有限，我們考慮照片間的內容以及彼此的關係，並且將版面組織的轉變為一個條件最佳化的問題。 與單張照片播放的幻燈秀比較，實驗結果顯示我們的方法更能帶給使用者愉悅的照片瀏覽經驗。This thesis presents a new medium, called tiling slideshow, to display photos in a tile-like manner, coordinating with the pace of background music. In contrast to conventional photo slideshow, multiple photos that have similar characteristics are well arranged and displayed at the same layout. Motivated by the guidelines of technical writing, each displaying layout is composed of a larger topic photo and several small-size supportive photos. Based on this idea, the proposed tiling slideshow system consists of three major components: image clustering, music analyzer, and layout organizer. Given the limited displaying space, we consider the context and relationship between photos and model the layout organization as a constrainted optimization problem. Experiments on real consumer photograph collections show that the novel displaying method gives users more pleasant browsing experience than the methods that focus only on single photograph display.1 Introduction ........................................... 1 1.1 Motivation ........................................... 1 1.2 Related Works ........................................ 3 1.3 The Proposed Solution ................................ 4 1.4 Thesis Organization .................................. 6 2 System Overview ........................................ 7 2.1 Essential Idea ....................................... 7 2.2 System Framework ..................................... 8 3 Visual Processing ...................................... 13 3.1 Photo Preprocess ..................................... 13 3.1.1 Orientation Correction ............................. 13 3.1.2 Underexposure/Overexposure Photo Detection ......... 15 3.1.3 Duplicate Photo Detection .......................... 15 3.1.4 Blur Photo Detection ............................... 16 3.2 Image Clustering ..................................... 19 3.2.1 Time-based Clustering .............................. 20 3.2.2 Content-based Clustering ........................... 22 3.3 Region of Interest Determination ..................... 27 3.3.1 Region of Interest ................................. 27 3.3.2 Bottom-up Attention Detection ...................... 28 3.3.3 Top-down Attention Detection ....................... 31 3.4 Summary .............................................. 32 4 Music Analysis ......................................... 33 4.1 Beat Detection ....................................... 33 4.2 Music Segmentation ................................... 37 4.3 Summary .............................................. 39 5 Tiling Slideshow Composition ........................... 40 5.1 Photo Importance ..................................... 42 5.1.1 Cluster-based Importance ........................... 42 5.1.2 Photo-based Importance ............................. 43 5.2 Cluster Selection .................................... 44 5.3 Tiling Frame Generation .............................. 44 5.4 Template Importance .................................. 45 5.5 Template Determination ............................... 46 5.6 Composition .......................................... 47 5.6.1 Region Selection ................................... 48 5.6.2 Implementation ..................................... 49 5.6.3 Discussion ......................................... 50 6 Experimental Results ................................... 51 6.1 The Photo Content Set ............................... 51 6.2 The Subjective User Evaluation ....................... 52 7 Conclusions and Future Work ............................ 56 7.1 Conclusions .......................................... 56 7.2 Future Work .......................................... 57 References ............................................... 581202777 bytesapplication/pdfen-US幻燈秀影像分群興趣區偵測SlideshowImage clusteringRegion of interest detectionthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/53978/1/ntu-95-R93922025-1.pdf