Qualitative visual interpretation of 3D hand gestures using motion parallax
Roberto Cipolla, Yasukazu Okamoto, Yoshinori Kuno · 1992
We present an eficient and gmrnetricdly intuitive Problems with this approach algorithm to reliably interpret the image velocities of moving objects in 31). rt is well known that in a small neighbourbood the image motion of points on a plane ia characterised by an afine transformation. We show th & the relative image motion of a nearby non-coplanar point and its projection on the plane is equivalent to mo-tion pamE1nr and that it is a sehble geometric cue to 3D shape and viewer/ot>ject motion. In particular we ahow how to interpret the motion parallax vector of a fourth point and the turf, diue~gence and defownalion compw nents of the affine transformation (defined by the Ehree points of the plane) in order to recover the projection of the axis of rotation of a moving object: the change in relative poaition of the object; the rotation about the say; the till of the surface and a one parameter family of solutions for the slant as a function of the magnitude af tllc rotation of the object. The latter is a manifmtation of the bas-relief ambiguity. These measurements, allhough representing an incomplete solution to structure from motion, can be relierhly extracted from 2 views evcn when perspective effects are small. We present a real-time example in which the 3D vie sun1 interpretat ion of hand gestures or a hand-held object is used as part of a man-machine interface. Thin is an alternative to the Polhemua coil instrumented Data-gfav~ commonly uspd in sensing manual gestures. I