應用基於消失點之影像校正於強健式人臉偵測

Abstract

我們提出一種新的基於消失點影像校正方法以提高一般監控式系統性能，雖然大多數現有的人臉資料庫和偵測技術都是基於相機和人臉在同個高度上的假設，但在實際狀況下，相機很有可能會被安裝在不同的高度上，因此這種差異通常會降低基於正面人臉影像學習演算法的效能。在這篇論文當中，我們提出影像轉換來校正人臉影像，因此在訓練過程中可以不必蒐集不同的人臉方向。此外為了達到更佳的偵測效果，我們還提出基於膚色分析來降低錯誤偵測的方法。在實驗結果中表明我們所提出的方法在人臉偵測中可以得到顯著的改善。

We proposed a novel face detection method based on the vanishing point of vertical lines image rectification to improve system performance in a common surveillance application. While most existing face datasets and detection techniques are based on the assumption that the camera has a similar height as the target faces, in practical situations the camera may be installed at different heights. Such discrepancy often degrades the detection performance of algorithms based on learning with frontal face orientation. In this thesis we propose a transformation to rectify face images such that it is not necessary to collect training data of different face orientations. Furthermore, a method for reducing false alarms by skin analysis is proposed for better performance. The experiments show prominent improvements in face detection performance can be obtained with the proposed methods.