Method, apparatus and computer program product for generating panorama images

Patent metadata

Field	Value
Publication number	US-9619863-B2
Application number	US-201314395962-A
Country	US
Kind code	B2
Filing date	Apr 5, 2013
Priority date	Apr 25, 2012
Publication date	Apr 11, 2017
Grant date	Apr 11, 2017

Abstract

In accordance with an example embodiment a method, apparatus and computer program product are provided. The method comprise assigning weights to at least one first feature and at least one second feature. The at least one first feature may be associated with a first image and the at least one second feature may be associated with a second image. The weights are assigned based on a distance of the at least one first feature and the at least one second feature from a central portion of the corresponding one of the first image and the second image. The method further includes registering the first image and the second image based at least on the assigned weights to determine transforms between the first image and the second image. The first image and the second image may be based on the determined transform.

First claim

The invention claimed is:

1. A method comprising:
assigning, by an image processor, weights to at least one first feature and at least one second feature, the at least one first feature being associated with a first image and the at least one second feature being associated with a second image, the weights being assigned based on a distance of the at least one first feature and the at least one second feature from a center of an overlapping portion of the first image and the second image, wherein the weights decrease exponentially from the center of the overlapping portion;
registering, by the image processor, the first image and the second image based at least on the assigned weights to determine transforms between the first image and the second image, wherein features closer to the center of the overlapping portion of the first image and the second image are preferentially used for determining the transforms; and
transforming, by the image processor, the first image and the second image based on the determined transform.

2. The method as claimed in claim 1 , wherein registering at least the first image and the second image comprises:
performing feature matching between the at least one first feature and the at least one second feature to generate a plurality of matching features; and
computing a plurality of matching weights for the plurality of matching features based on the weights assigned to the at least one first feature and the at least one second feature.

3. The method as claimed in claim 2 further comprising:
estimating a plurality of transforms between a set of matching features of the plurality of matching features, the set of matching features being selected from the plurality of matching features based on the plurality of matching weights; and
selecting a transform of the plurality of transforms based on a total matching weight associated with the transform, the total matching weight of the transform being determined based on a plurality of inliers associated with the transform.

4. The method as claimed in claim 1 , further comprising computing the at least one first feature and the at least one second feature prior to assigning weights to the at least one first feature and the at least one second feature.

5. The method as claimed in claim 1 , further comprising determining a distance of the at least one first feature from a central portion of the first image, and determining a distance of the at least one second feature from a central portion of the second image, prior to assigning weights to the at least one first feature and the at least one second feature.

6. The method as claimed in claim 3 , wherein the set of matching features is selected from the plurality of matching features by:
sorting the plurality of matching features in an order of the matching weights associated with each of the plurality of matching features; and
selecting the matching features of the plurality of matching features that have a matching weight greater than a predetermined matching weight as the set of matching features.

7. The method as claimed in claim 6 , wherein estimating the plurality of transforms from the set of matching features comprises:
selecting random pairwise matching features from the set of matching features;
computing the total matching weight for each of the selected random pairwise matching features; and
determining a transform corresponding to a maximum total matching weight from the computed total matching weight, the determined transform being utilized for panorama generation.

8. The method as claimed in claim 6 , wherein the total matching weight is determined by:
computing a probable transform for each of the random pairwise matching features;
determining an inlier corresponding to each of the probable transform; and
computing the total matching weight associated with each of the probable transforms for the determined inlier.

9. An apparatus comprising:
at least one processor; and
at least one memory comprising computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to at least perform:
assign weights to at least one first feature and at least one second feature, the at least one first feature being associated with a first image and the at least one second feature being associated with a second image, the weights being assigned based on a distance of the at least one first feature and the at least one second feature from a center of an overlapping portion of the first image and the second image, wherein the weights decrease exponentially from the center of the overlapping portion;
register the first image and the second image based at least on the assigned weights to determine transforms between the first image and the second image, wherein features closer to the center of the overlapping portion of the first image and the second image are preferentially used for determining the transforms; and
transform the first image and the second image based on the determined transform.

10. The apparatus as claimed in claim 9 , wherein, to register at least the first image and the second image, the apparatus is further caused, at least in part to:
perform feature matching between the at least one first feature and the at least one second feature to generate a plurality of matching features; and
compute a plurality of matching weights for the plurality of matching features based on the weights assigned to the at least one first feature and the at least one second feature.

11. The apparatus as claimed in claim 10 , wherein the apparatus is further caused, at least in part to perform:
estimate a plurality of transforms between a set of matching features of the plurality of matching features, the set of matching features being selected from the plurality of matching features based on the plurality of matching weights; and
select a transform of the plurality of transforms based on a total matching weight associated with the transform, the total matching weight of the transform being determined based on a plurality of inliers associated with the transform.

12. The apparatus as claimed in claim 9 , wherein the apparatus is further caused, at least in part to compute the at least one first feature and the at least one second feature prior to assigning weights to the at least one first feature and the at least one second feature.

13. The apparatus as claimed in claim 9 , wherein the apparatus is further caused, at least in part to determine a distance of the at least one first feature from a central portion of the first image, and determine a distance of the at least one second feature from a central portion of the second image, prior to assigning weights to the at least one first feature and the at least one second feature.

14. The apparatus as claimed in claim 11 , wherein, to select the set of matching features from the plurality of matching features, the apparatus is further caused, at least in part to:
sort the plurality of matching features in an order of the matching weights associated with each of the plurality of matching features; and
select the matching features of the plurality of matching features that have a matching weight greater than a predetermined matching weight as the set of matching features.

15. The apparatus as claimed in claim 14 , wherein, to estimate the plurality of transforms from the set of matching features, the apparatus is further caused, at least in part to:
select random pairwise matching features from the set of matching features;
compute the total matching wei

Assignees

• Nokia Technologies Oy

Inventors

• Muninder Veldandi
• S V Basavaraja

Classifications

• H04N23/698

  for achieving an enlarged field of view, e.g. panoramic image capture · CPC title

• G06T2200/32

  involving image mosaicing · CPC title

• G06T3/4038 Primary

  Image mosaicing, e.g. composing plane images from plane sub-images · CPC title

• H04N5/23238

  Electricity · mapped topic

• G06K9/4609

  Physics · mapped topic