Issue No. 07 - July (1993 vol. 15)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/34.221167
<p>Image flow fields from parallel stereo cameras are analyzed to determine the relative 3-D translational motion of the camera platform with respect to objects in view and to establish stereo correspondence of features in the left and right images. A two-step procedure is suggested. In the first step, translational motion parameters are determined from linear equations the coefficients of which consist of the sums of measured quantities in the two images. Separate equations are developed for cases when measurements of either the full optical flow or the normal flow are available. This computation does not require feature-to-feature correspondence. In addition, no assumption is made about the surfaces being viewed. In the second step of the calculation, with the knowledge of the estimated translational motion parameters, the binocular flow information is used to find features in one image that correspond to given features in the other image. Experimental results with synthetic and laboratory images indicate that the method provides accurate results even in the presence of noise.</p>
binocular image flows; parallel stereo cameras; relative 3-D translational motion; translational motion parameters; linear equations; optical flow; normal flow; binocular flow; computer vision; motion estimation; optical information processing
J. Duncan and L. Li, "3-D Translational Motion and Structure from Binocular Image Flows," in IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 15, no. , pp. 657-667, 1993.