System and method for scene dependent multi-band blending
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06K-009/40
G06T-005/00
출원번호
US-0331260
(2014-07-15)
등록번호
US-9454803
(2016-09-27)
발명자
/ 주소
Hsu, Stephen Charles
출원인 / 주소
Google Inc.
대리인 / 주소
Dority & Manning, P.A.
인용정보
피인용 횟수 :
0인용 특허 :
6
초록▼
Systems and methods for generating a composite image from a plurality of source images using a scene dependent multi-band blending operation are provided. The multi-band blending operation implements a filtering operation to reduce blending between objects or surfaces that have natural color and/or
Systems and methods for generating a composite image from a plurality of source images using a scene dependent multi-band blending operation are provided. The multi-band blending operation implements a filtering operation to reduce blending between objects or surfaces that have natural color and/or brightness differences. More particularly, the typical space invariant upsampling that occurs during multi-band blending can be replaced by a scene dependent filtering operation during upsampling that omits or reduces contributions from pixels associated with different objects in a scene during the multi-band blending process. The scene dependent filtering can be based on scene dependent data, such as height data or slope data, which can be used to identify different objects in a scene.
대표청구항▼
1. A computer-implemented method of blending a plurality of source images, comprising: obtaining, by one or more computing devices, a plurality of source images depicting a scene, each of the plurality of source images comprising a plurality of pixels;accessing, by the one or more computing devices,
1. A computer-implemented method of blending a plurality of source images, comprising: obtaining, by one or more computing devices, a plurality of source images depicting a scene, each of the plurality of source images comprising a plurality of pixels;accessing, by the one or more computing devices, scene-dependent data associated with the plurality of source images, the scene dependent data indicative of different objects in the scene; andperforming, by the one or more computing devices, a multi-band blending operation to generate a composite image from the plurality of source images, the multi-band blending operation comprising a filtering operation based on the scene dependent data to modify blending of pixels located at a boundary between the plurality of source images. 2. The computer-implemented method of claim 1, wherein the scene dependent data comprises elevation data, slope data, or object identity data. 3. The computer-implemented method of claim 1, wherein the filtering operation reduces multi-band blending between pixels associated with objects of different elevation located at the boundary between the plurality of source images. 4. The computer-implemented method of claim 1, wherein the filtering operation reduces multi-band blending between one or more pixels associated with a first object depicted in a first source image and one or more pixels associated with a second object depicted in a second source image. 5. The computer-implemented method of claim 1, wherein the source images are geographic images depicting a geographic area. 6. The computer-implemented method of claim 5, wherein the composite image is texture mapped to a three-dimensional surface to provide a three-dimensional representation of the geographic area. 7. The computer-implemented method of claim 1, wherein performing, by the one or more computing devices, the multi-band blending operation to generate the composite image comprises: constructing, by the one or more computing devices, a first Gaussian pyramid for a first source image and a second Gaussian pyramid for at least one second source image;converting, by the one or more computing devices, the first and second Gaussian pyramids into first and second Laplacian pyramids;generating, by the one or more computing devices, a composite Laplacian pyramid from the first and second Laplacian pyramids;reconstructing, by the one or more computing devices, a composite Gaussian pyramid from the composite Laplacian pyramid, the composite Gaussian pyramid having a plurality of levels including a base level and a plurality of higher levels associated with progressively lower levels of resolution, each level having one or more Gaussian pixels having a reconstructed Gaussian pixel value; andextracting, by the one or more computing devices, the composite image from the base level of the composite Gaussian pyramid. 8. The computer-implemented method of claim 7, wherein the filtering operation is performed during reconstruction of the composite Gaussian pyramid. 9. The computer-implemented method of claim 8, wherein the filtering operation reduces the combining of pixel values associated with objects of different elevation during calculation of reconstructed Gaussian pixel values for Gaussian pixels in the composite Gaussian pyramid. 10. The computer-implemented method of claim 8, wherein the filtering operation reduces the combining of pixel values associated with different objects depicted in the scene during calculation of reconstructed Gaussian pixel values for Gaussian pixels in the composite Gaussian pyramid. 11. The computer-implemented method of claim 7, wherein the multi-band blending operation is implemented using a dual pyramid data structure, the dual pyramid data structure having a plurality of nodes, one or more of the plurality of nodes having a payload providing a first Gaussian pixel value computed from the first source image and a second Gaussian pixel value computed from the second source image. 12. The computer-implemented method of claim 11, wherein the filtering operation uses, based on the scene dependent data, one of the first Gaussian pixel value or the second Gaussian pixel value in computing a reconstructed Gaussian pixel value for a Gaussian pixel in the composite Gaussian pyramid. 13. A computing system configured to generate a composite image from a plurality of source images, the computing device having one or more processors and one or more memory devices, the one or more memory devices storing non-transitory computer-readable instructions for execution by the one or more processors to cause the one or more processors to perform operations, the operations comprising: obtaining a plurality of source images depicting a scene, each of the plurality of source images comprising a plurality of pixels;accessing scene-dependent data associated with the plurality of source images, the scene dependent data indicative of different objects in the scene; andperforming a multi-band blending operation to generate a composite image from the plurality of source images, the multi-band blending operation comprising a filtering operation based on the scene dependent data to modify blending of pixels located at a boundary between the plurality of source images. 14. One or more tangible non-transitory computer-readable media storing computer readable instructions that when executed by one or more processors cause the one or more processors to perform operations, the operations comprising: obtaining a plurality of source images depicting a scene, each of the plurality of source images comprising a plurality of pixels;accessing scene-dependent data associated with the plurality of source images, the scene dependent data indicative of different objects in the scene; andperforming a multi-band blending operation to generate a composite image from the plurality of source images, the multi-band blending operation comprising a filtering operation based on the scene dependent data to modify blending of pixels located at a boundary between the plurality of source images.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (6)
Yalin Xiong, Blending arbitrary overlaying images into panoramas.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.