Camera Recognition
暫譯: 相機識別
nist
- 出版商: CreateSpace Independ
- 出版日期: 2013-11-13
- 售價: $810
- 貴賓價: 9.5 折 $770
- 語言: 英文
- 頁數: 64
- 裝訂: Paperback
- ISBN: 1493758950
- ISBN-13: 9781493758951
海外代購書籍(需單獨結帳)
商品描述
The Department of Homeland Security’s (DHS) United States Visitor and Immigrant Status Indicator Technology (US-VISIT) program is a biometrically-enhanced identification system primarily situated at border points of entry such as airports and seaports. In a 2004 assessment of the quality of facial images captured by US-VISIT, the National Institute of Standards and Technology (NIST) discovered a widespread problem: many subjects were 1) not directly facing the camera and 2) had a pose angle of greater than 10 degrees. The findings of NIST’s subsequent follow-up studies suggest that the camera used to capture facial images of travelers should look as much as possible like a traditional camera. Knowing where to look will help the subjects being photographed orient themselves in such a way that they are frontal to the camera – thus improving picture quality. This study explored whether participants could discern image capture devices (i.e., cameras) from other types of technology, and the attributes they relied upon to make that distinction.
商品描述(中文翻譯)
美國國土安全部(DHS)的美國訪客與移民身份指標技術(US-VISIT)計畫是一個以生物識別技術增強的識別系統,主要設置在邊境入境點,如機場和海港。在2004年對US-VISIT所捕捉的臉部影像質量的評估中,美國國家標準與技術研究所(NIST)發現了一個普遍的問題:許多受試者1) 並未正對著相機,且2) 姿勢角度超過10度。NIST隨後的後續研究結果顯示,用於捕捉旅客臉部影像的相機應儘可能看起來像傳統相機。知道該往哪裡看將幫助被拍攝的受試者以正面朝向相機的方式進行定位,從而改善影像質量。本研究探討參與者是否能夠辨別影像捕捉設備(即相機)與其他類型的技術,以及他們依賴哪些特徵來做出這種區分。