3d object motion from 3d object shape

Estimation of 3D object motion two succeeding video images with known object shape

Goal

This demonstration is on automatic analysis of moving objects in a video sequence. The goal is the estimation of the 3D motion of a 3D object.

Approach

The 3D object motion is estimated using the following assumptions:

• 3D object shape is known
• 2D object segmentation mask is known
• Object is rigid
• Scene illumination is diffuse
• Camera can be approximated by orthogonal projection

The object motion is estimated by an optical flow approach. Using the known 3D shape, 3D motion parameters are introduced in the optical flow equations [LeSaux].

Experimental Results

The experimental results are derived as follows:

• Object segmentation: By COST 211 Analysis Model [COST]
• 3D shape estimation: Shape is derived by a simple distance transform from the 2D segementation mask
• 3D motion estimation: For each image w.r.t. the preceding image

Original image sequence Object segmentation mask 3D object shape Estimated projected rotation axis

The results vizualise the projection of the 3D object rotation axis into the 2D image plane. The estimated motion follows well the motion of the object. Some temporal jitter is due to the independent estimation for each image. The results show that the applied assumptions are valid for simple video scenes.

References

[COST]

COST 211quat Analysis Model

[LeSaux99]

B. Le Saux: "Estimation du mouvement d'un objet avec un modèle de caméra simplifié", rapport de stage de DEA, IRISA/INRIA, Rennes, France, Septembre 1999.

Up to motion page