REAL TIME VIDEO STABILIZATION
First Claim
1. A method for video stabilization of a sequence of images captured by a camera, the sequence of images comprising at least one successive image pair with a current image, Ii+1, and a previous image, Ii, i≧
- 0 denoting an image frame number, and the method comprising the steps of;
estimating an optical flow from the previous image, Ii, to the current image, Ii+1;
determining an Euclidean transformation, Fi, from the previous image, Ii, to the current image, Ii+1, wherein the Euclidean transformation, Fi, comprises a product of at least a rotation matrix, Ri, and a scaling factor, si, and a translation matrix, Ti;
restricting the transformation of the current image, Ii+1, by using Euclidean matrices Ai and Bi, wherein;
Ai is used to store absolute frame transformation relative world frame coordinates, A0 is the identity matrix and Ai+1 is equal to AiFi−
1 for i≧
0 and Bi is used to store visualization window coordinate transformation;
transforming the current image, Ii+1, by applying the inverse of the Euclidean transformation, Fi, to the current image, Ii+1.
1 Assignment
0 Petitions
Accused Products
Abstract
The present invention provides a method and a system for video stabilization in real time. The method comprises the steps of: estimating an optical flow from an previous image, Ii, to a current image, Ii+1; determining an Euclidean transformation, Fi, from the previous image, Ii to the current image, Ii+1, wherein the Euclidean transformation, Fi, comprises a product of rotation, Ri, and scaling, si, and translation, Ti; and transforming the current image, Ii+1, by applying the inverse of the Euclidean transformation, Fi, to the current image, Ii+1 in order to obtain stabilization. The transformation of the current image, Ii+1, is restricted by using Euclidean matrices Ai and Bi, wherein: Ai is used to store absolute frame transformation relative world frame coordinates, A0 is the identity matrix and Ai+1 is equal to AiFi−1 for i≧0 and Bi is used to store visualization window coordinate transformation.
11 Citations
9 Claims
-
1. A method for video stabilization of a sequence of images captured by a camera, the sequence of images comprising at least one successive image pair with a current image, Ii+1, and a previous image, Ii, i≧
- 0 denoting an image frame number, and the method comprising the steps of;
estimating an optical flow from the previous image, Ii, to the current image, Ii+1; determining an Euclidean transformation, Fi, from the previous image, Ii, to the current image, Ii+1, wherein the Euclidean transformation, Fi, comprises a product of at least a rotation matrix, Ri, and a scaling factor, si, and a translation matrix, Ti; restricting the transformation of the current image, Ii+1, by using Euclidean matrices Ai and Bi, wherein;
Ai is used to store absolute frame transformation relative world frame coordinates, A0 is the identity matrix and Ai+1 is equal to AiFi−
1 for i≧
0 and Bi is used to store visualization window coordinate transformation;transforming the current image, Ii+1, by applying the inverse of the Euclidean transformation, Fi, to the current image, Ii+1. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
- 0 denoting an image frame number, and the method comprising the steps of;
-
9. A system for video stabilization of a sequence of images captured by a camera, the sequence of images comprising at least one successive image pair with a current image, Ii+1, and a previous image, Ii, i≧
- 0 denoting an image frame number, wherein the system comprises;
an estimation module adapted to for estimating the optical flow from the previous image, Ii, to the current image, Ii+1, a calculation module adapted to determine an Euclidean transformation matrix, Fi, of order three from the previous image, Ii, to the current image, Ii+1; a transformation module adapted to transform the current image, Ii+1 by applying the Euclidean transformation matrix, to obtain stabilization.
- 0 denoting an image frame number, wherein the system comprises;
Specification