Lab 공간에서의 방향벡터기반 IMM을 이용한 야간환경에 강인한 보행자 검출 알고리즘 연구

다국어 초록 (Multilingual Abstract)

This paper proposes pre-processing method for tracking the night of pedestrians in real time. Firstly, the nighttime image data convert to L＊a＊b＊color space. In this process, it is possible to extract a robust L＊space about noise removal and brightness adjustment. By using L*space data and the video subtraction method, it makes pre-processing data. Since then it detects the optimal pedestrian silhouette using DV-IMMP probability algorithm which is used removed background image, proposed DV-filter and HP-SVM filter. After that it tracks the movement of pedestrians to use Kalman filter. Probability algorithm to be proposed can be using tracking of night pedestrians in real time and it also can replace Histograms of oriented gradients(HOG) and Haar-like algorithm that requires a lot of computational complexity.

참고문헌 (Reference)

1 김민재, "신체특징을 이용한 SVM필터링으로 관심지역에서의 수상한 보행자 검출" 한국정보기술학회 11 (11): 71-77, 2013

2 김동우, "멧돼지 감시 시스템을 위한 객체 검출 방법" 한국인터넷방송통신학회 10 (10): 229-235, 2010

3 이광국, "그림자 제거를 위한 색상 공간의 비교" 한국멀티미디어학회 11 (11): 610-622, 2008

4 이영학, "곡률과 HOG에 의한 연속 방법에 기반한 아다부스트 알고리즘을 이용한 보행자 인식" 한국정보과학회 16 (16): 654-662, 2010

5 Seung-Yeon Jeong, "Research for improving the detecting pedestrian with Night Vision" 33 (33): 1362-1364, 2010

6 Wang Bing-bing, "Pedestrian detection based on the combination of HOG and background subtraction method" 527-531, 2011

7 E. Mazor, "Interacting multiple model methods in target tracking: a survey" 34 : 103-123, 1998

8 Quan Pan, "Interacting multiple model algorithm based on α-β/α-β-γ filters" 6 : 3698-3702, 1997

9 I. P. Alonso, "Combination of Feature Extraction Methods for SVM Pedestrian Detection" 8 (8): 292-307, 2007

10 I. P. Alonso, "Combination of Feature Extraction Methods for SVM Pedestrian Detection" 8 : 292-307, 2007

11 Rainer Lienhart, "An Extended-Set of Haar-like Features for Rapid Object Detection" 1 : 900-903, 2002

12 Jong-dae Park, "An Advanced Method of Face Recognition Using Kernel-PCA" 8 (8): 470-472, 2010