結合多種特徵之近紅外線夜間行人偵測

Near-Infrared Based Nighttime Pedestrian Detection by Combining Multiple Features

林預淳(Yu-Chun Lin)

指導教授 ： 傅立成

共同指導教授 ： 蕭培墉(Pei-Yung Hsiao)

國立臺灣大學/電機資訊學院/資訊工程學研究所/碩士(2010年)

https://doi.org/10.6342/NTU.2010.01788

摘要

在電腦視覺的領域中，行人偵測是個重要的主題，而且其中夜間行人偵測更是特別的困難且有挑戰性。在這篇論文中，我們將偵測行人的問題定義於夜間的移動式平台使用相機所擷取的影像串流上。目前大多數的夜間行人偵測方法都只是使用圖像上的單一特徵為其核心。原本於日間的環境下使用很多有效的圖像特徵，換到夜間的環境時遭遇到了缺少材質、高對比以及低照度的問題。在這樣的問題條件下，我們首先使用智慧區域偵測方法切割出圖像中的前景以便產生出候選區域。在這之後我們設計一個夜間行人偵測系統基於AdaBoost以及支持向量機分類器，結合輪廓以及旋轉強度之方向長條圖特徵以達到有效的將候選區域辨識為行人與否。結合不一樣類型且互補的特徵可以增加偵測的效果。實驗結果展現出我們的行人偵測系統於夜間環境中達到所期望的成果。

關鍵字

夜間 ； 行人偵測 ； 輪廓 ； 方向梯度直方圖 ； 支持向量機

並列摘要

Pedestrian detection is an important subject in computer vision field, and the nighttime pedestrian detection is especially difficult and challenging. In this thesis, we address the problem of detecting pedestrians in video streams from a moving camera at nighttime. Most nighttime human detection approaches only use single feature extracted from images. The effective image features in daytime environment may suffer from textureless, high contrast and low light problems at night. To deal with these issues, we first segment the foreground by using the proposed Smart Region Detection approach to generate candidates. Then we design a nighttime pedestrian detection system based on the AdaBoost and support vector machine (SVM) classifiers with contour and histogram of oriented gradients (HOG) features to effectively recognize pedestrians from those candidates. Combining different type of complementary features can improve the detection performance. Experiment results show that our pedestrian detection system is promising in the nighttime environment.

並列關鍵字

nighttime ； pedestrian detection ； contour ； HOG ； SVM

參考文獻

[1] J. Li, "Intensity-Distance Projection Space Based Human Tracking in Far-Infrared Image Sequences," 2009, pp. 371-375.

[2] M. Bertozzi, et al., "Multi Stereo-Based Pedestrian Detection by Daylight and Far-Infrared Cameras," in Augmented Vision Perception in Infrared, ed, 2009, pp. 371-401.

[3] R. O'Malley, et al., "A review of automotive infrared pedestrian detection techniques," in Signals and Systems Conference, 208. (ISSC 2008). IET Irish, 2008, pp. 168-173.

[5] Q. M. Tian, et al., "Pedestrian detection in nighttime driving," in Third International Conference on Image and Graphics, Proceedings 2004, pp. 116-119.

[6] Y. Fang, et al., "Comparison between infrared-image-based and visible-image-based approaches for pedestrian detection," in IEEE Intelligent Vehicles Symposium, Proceedings. , 2003, pp. 505-510.