Moving Object Detection using Visible Spectrum Imaging and Thermal Imaging

摘要

This paper deals with development of Robust Moving Object Detection method using Visible Spectrum and Thermal Imaging. If we fuse visible spectrum and Thermal imaging together, more information about the moving object can be obtained as both are inherently complementary to each other. The segmentation of moving objects in corresponding video frames involves `Background Subtraction' method. Here each pixel of visible as well as thermal is modeled as mixture of Gaussians for background subtraction, then blob analysis and fusion rule is applied to build Robust moving Object detection system.