효과적인 ROI 탐지를 위한 시공간특징의 가중치 결정방법

다국어 초록 (Multilingual Abstract)

In this paper, a new weight decision method of spatiotempoal feature for efficient ROI(region of interest) detection is proposed. Most ROI detection system uses a cue that is integrated by multiple extracted spatiotemporal features from input image. It utilizes spatiotemporal features for the purpose of detecting ROI. In the past, it is assumed that motion representing temporal visual conspicuity between adjacent frames takes higher priority over spatial visual conspicuity. However if the movement of the object is very small, we usually pay attention to spatial visual conspicuity such as color or intensity, not the moving object. Therefore, proposed system sets the weight dynamically by relative activities of each of input features. To evaluate the performance of the system, we tested with various kinds of real video sequences. We also compared our system with several previous systems to validate the performance of the system.

참고문헌 (Reference)

1 노상수, "중요도 특징 지도 변화를 이용한 동영상 Visual Attention Model" 30 (30): 935-936, 2007

2 홍혜민, "인간의 네 가지 시각경로를 이용한 주의시각 기반 시각정보처리체계 시스템의 구현" 연세대학교 대학원 2006

3 류광근, "이동로봇의 물체인식을 위한 질의 기반 시각 집중 알고리즘" 대한전자공학회 44 (44): 50-58, 2007

4 이진영, "능동시각을 위한 모션정보 기반 선택적 주의 집중 시스템" 고려대학교 대학원 2008

5 Y. Zhai, "Visual Attention Detection in Video Sequences Using Spatiotemporal Cues"

6 M. Park, "Selective Visual Attention System Based on Spatiotemporal Features" 5068 : 203-212, 2008

7 Nitin Dhavale, "Saliency-based multi-foveated MPEG compression" 2003

8 J. Tsotsos, "Modeling Visual Attention via Selective Tuning" 78 : 507-545, 1995

9 D. Lowe, "Distinctive Image Features from Scale-Invariant Keypoints" 2004

10 S. Olivier, "Development of a Biologically Inspired Real-Time Visual Attention System" 1811 : 150-159, 2000

11 B.Choi, "Development of a Biologically Inspired Real-Time Spatiotemporal Visual Attention System" 6591 : 416-424, 2011

12 T. Hamamoto, "Column Parallel A/D Conversion on CMOS Image Sensor" 165-168, 2001

13 L. Itti, "A saliency-based search mechanism for overt and covert shifts of visual attention" 40 : 1489-1506, 2000