Method for taking panorama mosaic photograph with a portable terminal

• US 20080056612A1
• Filed: 06/14/2007
• Published: 03/06/2008
• Est. Priority Date: 09/04/2006
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd., Seoul National University Industry Foundation (Seoul National University)
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Kim, Soo-Kyun, Park, Kyoung-Ju, Cho, Nam-Ik, Ha, Seong-Jong, Lee, Sang-Hwa, Moon, Jae-Won, Koo, Hyung-Il, Cho, Sung-Dae
• Granted Patent: US 7,881,559 B2
• Time in Patent Office: Days
• US Class Current: 382/284
• CPC Class Codes: G06T 3/4038 Image mosaicing, e.g. composing plane images from plane sub-images
• CPC Class Codes: G06V 10/16 using multiple overlapping images; Image stitching
• CPC Class Codes: G06V 10/20 Image preprocessing

Abstract

A method for taking a panorama mosaic photograph includes displaying a partial image of a previously taken image as a guide image on a viewer of an image to be currently taken and taking a number of images constituting the panorama mosaic photograph according to a photography operation; projecting the taken images onto a common cylindrically curved surface; and joining the projected images into a single image.

Claims

1. A method for taking a panorama mosaic photograph, comprising:

(a) displaying a partial image of a previously taken image as a guide image on a display showing a current image to be taken and taking a number of images, in sequence, constituting the panorama mosaic photograph according to a photography operation;

(b) projecting the taken images onto a common cylindrically curved surface; and

(c) joining the projected images into a single image, wherein

step (b) is performed according to equation (12) below

2. The method as claimed in claim 1, wherein the radius R of the cylindrical coordinate is defined by equation (13) below

3. The method as claimed in claim 1, wherein, in step (a), the partial image of the previously taken image is transformed into a coordinate of the image to be currently viewed in the display.

4. The method as claimed in claim 1, wherein, in step (a), the previously taken image is projected onto the cylindrically curved surface and translated by the θ

, and an overlapping region between the previously taken image and the image to be currently taken is re-projected onto a planar surface and displayed as the guide image.

5. The method as claimed in claim 1, wherein, when the images are projected onto the cylindrically curved surface, a luminance difference is detected with regard to an overlapping region between the images so that the images are adjusted to have an identical luminance.

6. The method as claimed in claim 1, wherein step (c) comprises a aligning step, a stitching step, and a blending step.

7. The method as claimed in claim 6, wherein, in the aligning step, a hexagonal search technique is used to determine an alignment condition by obtaining a cost of an overlapping region between two images according to cost function |I_1.(x,y)^ov−

I_2.(x,y)^ov|<sub>(x,y)ε

overlapped area</sub> when two adjacent images I₁ and I₂ overlap each other in regions I₁^ov and I₂^ov and when an error surface is e=|I₁^ov−

I₂^ov|, moving I₁ or I₂ with regard to locations of six peripheral pixels at a predetermined distance from current location (0,0) so as to obtain a cost for each case, and finding a minimum cost, and,

in the stitching step, a dynamic program is used to obtain cumulative error E with regard to an overlapping region between two aligned images according to E(i, j)=e(i, j)+min(E(i−

1, j−

1), E(i, j−

1), E(i+1, j−

1)) and find a path having a smallest cumulative error.