作者: tenz

這台怪物據說是 Google Books Search 用來把書頁內容掃進資料庫以供搜尋的究極兵器。在這篇報導中介紹了這台怪兵器的厲害之處：

According an essay written by Joseph Huttner, “Google recently implemented a convoy of new machines from Kirtas Technologies called the APT BookScan 1200, which is capable of scanning up to 1200 pages per hour (Kirtas Technologies). The machine works by utilizing a 16-megapixel digital camera to photograph each page, then transfers the image to local storage. Then, a robotic arm gently turns the page of the book, and the process is completed for each additional page. The digital images are then ‘cleaned’ to remove smudges and other errors, cropped, and centered. At this point, the image can be posted online, but “searching through the text is impossible”.

文章雖說在拍完後仍無法對內文進行搜尋，但圖檔很快就會送到 Google 的 OCR 軟體去辨識成數位化的文字資料，這樣就可以搜尋了。BTW，Google 說最近會推出支援中文的 Google Book Search。到時有多少台灣和中國的出版社想跟 Google 玩呢？值得觀察。