Somayeh Kermani
Kamal Jamshidi
Amirhassan Monadjemi

3D motion detection requires, on the one hand, the study of motion of the objects in the environment, and on the other hand, the depth estimation of the scene. These two tasks are computationally expensive and are not suitable for real time application. This paper proposed a method for the detection of moving objects in the stereo image sequences from a moving platform. In order to reduce the stereo matching time, moving objects are detected firstly and then a stereo vision method that uses motion information to perform depth estimation of the moving object is performed. The experimental results on real image sequences demonstrate that the proposed method can reduce the time complexity of the stereo matching and depth estimation.

Motion Detection, Optical Flow, Stereo Matching