人臉偵測在生物識別研究中為重要議題之一，因為人臉偵測在大部分的生物識別系統與監控系統中為識別與追蹤前必需處理的步驟。有好的人臉偵測結果，才能提高這些應用系統的效能。本篇論文對人臉影像取得LBP (Local Binary Pattern) 紋理特徵，並且使用Adaboost來訓練測試樣本，找出具有良好區分人臉與非人臉之特徵點，得到由這些特徵點形成的分類器。在偵測過程中，對輸入影像做不同的大小縮放，對每張縮放影像掃描找出人臉可能存在的區域，最後結合每張縮放影像偵測結果，偵測出輸入影像中各種不同大小的人臉。使用本篇論文提出的一個人臉偵測方法，來偵測出正確的人臉位置。實驗結果顯示，本篇論文所提出的方法可以對不同角度、人種、大小、表情、髮型及戴眼鏡或有鬍子的人臉有不錯的偵測效果。 Human face detection is among the most important topics in biometric research since it has a broad range of applications. Detection of face is often performed prior to recognition and tracking in biometric and surveillance systems. This paper proposes a face detection method based on Adaboost algorithm with LBP features. LBP feature is used for local feature representation due to its high discriminative power for texture classification and its invariance to global intensity variations. The best features are trained by Adaboost to form a strong classifier for distinguishing face and non-face images. In the test phrase, the input image is scaled with a number of factors to introduce a number of images of different sizes for face detection, so that faces of different sizes can be detected. Experimental results show that our method can detect faces with slight rotation, different skin colors, different facial expression, glasses and mustache.
