http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Thuong Khanh Tran,Ngoc Nam Bui,Jin Young Kim(김진영) 한국정보기술학회 2015 한국정보기술학회논문지 Vol.13 No.1
Human detection is a challenging problem in video processing, which is applied in many fields: robot control, surveillance system, traffic tracking etc. Recently, there have been many publications involving this problem. However, most of methods still focus on pedestrian detection. In this paper, based on the poselet techniques, we introduce a new method to detect human in video under various environments. By combining poselet and gradient local auto-correlation classifier, we propose an efficient technique in human detection and reduce false detection. Also, focused on edge-based robust principal component analysis, a new foreground extraction method is developed to handle the ambiguous environment such as: leaf motion, illumination etc. By applying the proposed method, the small motion artifacts can be rejected. Experimental results show that our method has the high accuracy in various environments.
Ngoc Nam Bui,Thuong Khanh Tran,So Hee Min(민소희),Jin Young Kim(김진영) 한국정보기술학회 2015 한국정보기술학회논문지 Vol.13 No.3
Human Action Recognition (HAR), in recent years, has attracted much attention from the research community due to its challenges as well as wide applications. In this paper, we investigate Universal Background Model (UBM) based GMM supervector and Support Vector Machine (SVM) with dense trajectories and motion bound features for HAR system. A GMM supervector is obtained by MAP adaptation with UBM and cascading all the mean vector components. After that, supervectors are applied as input features to SVM classifier with several kernels including modified non-linear GMM KL and GUMI kernels. Moreover, we also adopted channel fusion that used to enhance the robustness of classify model. Then we make a comparison and critical analysis between our method with those existing systems. Experimental results demonstrates that the proposed approach performs more efficient than current systems.