人工智慧及擴增實境的自拍系統｜國立臺灣科技大學博碩士論文系統

簡易檢索 / 詳目顯示

回結果列表

研究生：	蕭春元 Chun-Yuan Hsiao
論文名稱：	人工智慧及擴增實境的自拍系統 Multi-person Selfie System by Augmented Reality
指導教授：	楊傳凱 Chuan-Kai Yang
口試委員:	孫沛立 Pei-Li Sun 花凱龍 Kai-Lung Hua
學位類別：	碩士 Master
系所名稱：	管理學院 - 資訊管理系 Department of Information Management
論文出版年：	2018
畢業學年度：	106
語文別：	中文
論文頁數：	71
中文關鍵詞：	人像識別、擴增實境、自拍、影像接合、APP
外文關鍵詞：	Human recognition, augmented reality, selfie, image stitching, APP
相關次數：	點閱：884 下載：36
分享至:	分享至facebook 分享至twitter

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

本論文要完成的目的是製做雙人自拍系統APP，處理流程會分為兩大步驟，透過兩段式的拍攝方式，以完成合照。第一步驟是由使用者A幫使用者B先拍攝第一張照片，然後App會以人工智慧的方式將人像從照片中切割出來。第二個步驟則是反過來，由使用者B幫使用者A拍攝第二張照片，拍攝時APP 會由攝影機影鏡頭取得即時影像,並將第一步驟中切割出來的人像，透過擴增實境的方式將人像與即時影像做結合，最後再透過圖片的接合技術將兩張照片合而為一。

The purpose of this paper is to build a selfie system app for taking a multi-person photo. The process is takes two steps to complete the photo. The first step is for user A to take the first photo of user B, and then the App will cut the “portrait B” out of the photo in an artificially intelligent manner. The second step is for B to take the second photo of user B. During the shooting, the APP will acquire instant images from the camera lens. In addition, the APP will automatically calculate the relative position and proportion of “Portrait B” in the second photo. Then, through augmented reality, the “portrait B” is combined with the real-time image, and directly presented on the screen in a timely manner. In this way, users can look ahead and see if the final photo output is satisfactory. Finally, the two photos are combined into one by an image stitching technique.

中文摘要    I
Abstract    II
誌 謝    III
目  錄    i
圖目錄    iii
表目錄    v
第一章    緒論    1
1.1.    研究背景與動機    1
1.2.    研究問題    3
1.3.    研究目的    3
1.4.    論文架構    4
第二章    文獻探討    5
2.1.    人像偵測    5
2.2.    影像接合    9
2.3.    擴增實境    10
第三章    研究設計    12
3.1.    研究範圍    12
3.2.    研究流程與步驟    12
3.3.    系統設計    12
3.4.    人像偵測演算法試驗    17
3.5.    影像接合演算法試驗    32
第四章    系統實作    42
4.1.    環境建置    42
4.2.    影像接合系統實作    44
4.3.    擴增實境系統實作    46
4.4.    人像偵測系統實作    48
4.5.    系統整合    49
第五章    結論    54
5.1.    使用者試用感想    54
5.2.    軟體比較    55
5.3.    結論及檢討    58
5.4.    未來展望    58
第六章    參考文獻    60

                                

[1] N. Dalal, B. Triggs. (2005). Histograms of oriented gradients for human detection. IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
[2] Kaiming He, Georgia Gkioxari, Piotr Dollár, Ross Girshick. (2018). Mask R-CNN. arXiv:1703.06870v3 [cs.CV] 24 Jan 2018
[3] Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun;(2016). Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770-778
[4] David G. Lowe, Object Recognition from Local Scale-Invariant Features. Computer Science Department University of British Columbia
[5] Herbert Bay, Tinne Tuytelaars, Luc Van Gool, SURF: Speeded Up Robust Features.
[6] R. Azuma. A Survey of Augmented Reality Presence: Teleoperators and Virtual Environments
[7] P. Milgram, A. F. Kishino. Taxonomy of Mixed Reality Visual Displays.
[8] Shanshan Zhang, Christian Bauckhage, Armin B. Cremers. (2014). Informed Haar-like Features Improve Pedestrian Detection. CVPR2014, Computer Vision Foundation
[9] Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun. (2017). Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. IEEE Transactionson Pattern Analysis & Machine Intelligence 2017 vol.39 pp.1137-1149
[10] B. Triggs, P. McLauchlan, R. Hartley and A. Fitzgibbon. (1999). Bundle Adjustment — A Modern Synthesis, Vision Algorithms: Theory and Practice.
[11] Wikipedia. https://en.wikipedia.org/wiki/Max-flow_min-cut_theorem
[12] Wikipedia. https://en.wikipedia.org/wiki/Augmented_reality
[13] Wikipedia. https://en.wikipedia.org/wiki/Haar-like_feature
[14] Paul Viola Michael Jones,(2001) Robust Real-time Object Detectio
[15] Cortes C, Vapnik V,(1995) Support-vector networks. Machine Learning.

簡易檢索 / 詳目顯示

相關論文