Image processing method with motion estimation and image processing arrangement
View Patent ↗A method for processing a video image sequence with a number of successive images and an image processing device for processing a video image sequence are described. The method comprises: generation of at least one scaled image sequence (F 11 , F 12 ) from the video image sequence (F); generation of a motion-compensated scaled image sequence (F 40 ) from the scaled image sequence (F 50 ); generation of a motion-compensated output image sequence (F′) from the scaled image sequence (F 40 ); wherein the generation of the motion-compensated scaled image sequence (F 40 ) comprises the determination of a motion vector field based on at least two images (F 11 n−1 ; F 11 n ) of the scaled image sequence (F 11 , F 12 ) and the interpolation of an image using the at least two images and the motion vector field, and wherein the at least one input scaling factor (S 11 , S 12 ) is dependent on motion information that is represented by at least one motion vector field that was determined for the temporally preceding images of the at least one scaled image sequence (F 11 , F 12 ).
1. A method for processing a video image sequence, comprising:
generating an input scaling factor based on a speed of movement between corresponding image sections respectively located in current and preceding image frames;
scaling a size of an image section of the current image frame with the input scaling factor to produce a scaled image;
motion compensating the scaled image to produce a motion compensated scaled image;
scaling the motion compensated scaled image with an output scaling factor that is an inverse of the input scaling factor to produce a motion-compensated output image.
2. The method according to claim 1 , wherein the input scaling factor comprises a horizontal scaling component for scaling in a horizontal direction of an image and a vertical scaling component for scaling in a vertical direction of an image.
3. The method according to claim 2 , wherein the horizontal and vertical scaling components are different.
4. The method according to claim 3 , wherein each of the horizontal and vertical scaling components of the input scaling factor lies between 1 and 0.2.
5. The method of claim 1 wherein generating the input scaling factor comprises generating the input scaling factor based on a length of at least one motion vector representing movements between the image sections.
6. The method of claim 5 wherein generating the input scaling factor comprises generating the input scaling factor based on a longest motion vector of a motion vector field representing movements between the image sections.
7. The method of claim 1 wherein generating the input scaling factor comprises generating the input scaling factor based on an average length of motion vectors of a motion vector field of the image section in the current image frame.