基於被動式生活紀錄器之自動旅遊部落格產生系統

是否曾經有過，在旅遊中因為拍照花費太多時間，而錯失了走訪其他景點的機會，造成旅遊的遺憾。是否曾經有過拿起手機及相機時，想拍的景色已經改變。本論文被動式生活紀錄器以照片呈現最真實的生活記錄，純粹照相功能讓使用者能排除一切外在干擾。只需將裝置設定好後掛於胸前，便可依據設定的時間拍攝照片，捕捉生活中每個平凡而重要的片刻，讓使用者在旅途中可以盡情的玩樂，同時也可藉由被動式生活紀錄器記錄旅遊的美好。根據研究指出，超過九成的受訪者有經營或瀏覽他人部落格的經驗，內容方面以「旅遊札記」最為讀者所青睞，反應出目前國內民眾對旅遊休閒相關訊息的渴望與重視。但有些愛好旅遊的民眾，有時會因為家庭因素以及工作壓力，導致沒有多餘的時間，能夠專心的寫旅遊部落格與好友分享。本論文實作一套旅遊部落格產生系統，使用被動式生活紀錄器紀錄的影像，在短時間內，製作一篇圖文並茂的旅遊部落格。論文中使用以手機平台為基礎的「被動式生活紀錄器」及開發一套 PC端的「旅遊部落格產生系統」。前者主要為拍攝生活中的影像資訊，每張拍攝的影像中會記錄使用者拍攝當下的GPS位置，讓使用者可在手機中以地圖的方式瀏覽曾經去過的地點。後者的系統先將影像分割成事件，事件的分割依據每張影像紀錄的GPS 數據，經由公式，換算成速度，再依據速度的變化分割事件，接著會由事件中透過人臉偵測選出關鍵影像，並根據時間資訊對應到外出旅遊前所建立的Google行事曆內容，自動產生一篇旅遊部落格。

關鍵字

被動式生活紀錄器；旅遊部落格產生系統；事件分割；人臉偵測；關鍵影像選擇

並列摘要

Did you ever experience that spend too much time to take a picture, and miss other sights? Did you ever want to take a picture with the scenery you want to take and it changed when you taking a picture? This thesis proposes to use a passive life logging device to record your life and show the reality of your life time. Follow the steps then you can catch every moment in your daily life. First, you should set the device then hang on the chest. Then, the device will work and take pictures every fixed time interval. Having this device you can enjoy your time and have more fun. According to the study over 90% of the respondents have the experience to be a blogger or browsing others blogs. Traveling diaries is the most popular content in blogs. Considering some people cannot have enough time to post a new article and share the memories about wonderful time, this study will set an automatic traveling blog system. This system is using a passive life logging device to collect pictures and post new article in your blog. The thesis uses "Passive Life Logging Device" and puts forth "Automatic Travel Blogger System". The former’s function is mainly to take pictures of daily life and the GPS will record every image’s location to provide user to check them. The latter is to divide those images into events based on GPS data record of images. By the formula, GPS data is converted into speed, and then image data are segmented based on the speed diversification. Through face detection, a keyframe is selected from every event. According to the time correspondence of the travel schedule on Google calendar before going out, a travel blog is automatically produced from the keyframes.