A motion region estimation technique for Web camera application

S.Y. Koay1, A.R. Ramli1, Y.P. Lew1, V. Prakash1, R. Ali1
1Multimedia and lnrelligent System Research Group, Department of Compuler and Communication System Engineering. Faculty of Engineering, Universita Putra Malaysia, Selangor Darul Ehsan, Malaysia

Tóm tắt

This paper presents the steps involved in processing the output image obtained from image subtraction. A Web camera or Web cam is used as a video clip capture device and MATLAB version 6.01 with Image Processing Toolbox is used as the analysis software. Image frames extracted from the video clips undergo image subtraction for motion detection purposes. The continuous steps presented in this paper are thresholding, noise removal and motion region estimation. Thresholding determines the areas of output images (from subtraction) consisting of pixels with values lying within the threshold value range. Threshold value also indicates the sensitivity of motion to detection. The image still contains a small amount of noise after thresholding. Noise is removed using a median filtering method. Motion region estimation is done by executing an AND or OR operation on two subtracted images from three successive frames. Both the output image from AND and OR operations will estimate the motion region in a different time frame.

Từ khóa

#Motion estimation #Cameras #Motion detection #Application software #MATLAB #Image processing #Image analysis #Software tools #Pixel #Filtering

Tài liệu tham khảo

kameda, 1996, A Human Motion Estimation Method Using 3-Successive Video Frames, Proceedings of International Conference on Virtual Systems and MultiMedia, 135 10.1006/cviu.2000.0897 koay, 2002, A Motion Detection Technique for Web Camera Application, World Engineering Congress Proc liyama, 2000, Estimation of the Location of Joint Points of Human Body from Successive Volume Data, IAPR International Conference on Pattern Recognition, 3, 699 1997, Image Processing Toolbox User's Guide meyer, 1997, Model Based Extraction of Articulated Objects in Image Sequences, Fourth int conf on Image Processing, 10.1109/ICIP.1997.631988 10.1109/CVPR.1993.341008