Patent №
US 6,804,419
Granted
2004-10-12
Filed 1999
Owner
CANON KABUSHIKI KAISHA
Lab
—
AI components
1
vision
Assignment
Recorded
Dataset
AIPD
2023_r1 edition
Application
09434378
There is disclosed a method for obtaining one high-resolution image from a plurality of low-resolution images having predetermined resolutions, comprising steps of detecting relative positions among the plurality of low-resolution images with a resolution less than a pixel pitch in the predetermined resolutions, and forming a new image having a high resolution as compared with the predetermined resolutions using the plurality of images in accordance with information indicating the relative positions, so that the high-resolution image is obtained without any disorder. Moreover, there is disclosed a method for detecting motion vectors among a plurality of frames with a higher resolution, comprising steps of using orthogonal transform coefficients of images in the plurality of frames in motion images to detect the motion vectors among the plurality of frames with the resolution less than the pixel pitch in the predetermined resolutions.
AI classification
Ownership
CANON KABUSHIKI KAISHA
assignment · 105830140
Assignors
MIYAKE, NOBUTAKA
On an employer assignment, the assignors are typically the inventors.