Back to results
US 7,598,977 B2Granted
Spatio-temporal graphical user interface for querying videos
AssigneeMitsubishi Electric Research Laboratories, Inc.
InventorsKathleen Ryall, Qing Li, Alan W. Esenther
Life of this patent
Estimating the term…
Abstract
A method summarizes a video acquired of a scene spatially and temporally by specifying a spatial query and a temporal query. Frames of the video are selected according to the spatial query and the temporal query, and then the selected frames are composited into an image that represents spatial and temporal characteristics of the scene.
Claims20
A method for summarizing a video spatially and temporally, comprising:
- specifying a spatial query for a video acquired of a scene by a camera, in which the spatial query specifies a spatially contiguous set of pixels within a particular frame of the video;
- specifying a temporal query for the video, in which the temporal query specifies a temporally contiguous set of pixels in multiple adjacent frames;
- selecting frames of the a video according to the spatial query and the temporal query;
- and
- compositing the selected frames into an image that represents spatial and temporal characteristics of the scene, in which the compositing is according to a predetermined order and predetermined opacities, in which the compositing is according to:
- ∑
- i
- =
- 1
- n
-
- α
- i
- ×
- i
- n
- =
- 1
- ,
-
- and
-
-
- α
- i
- ×
- i
- n
- =
- α
- j
- ×
- j
- n
-
-
- for
-
-
- all
-
-
- i
- ≠
- j
- ,
- where αi is an opacity for an ith layer in the image, and n represents a number of selected images.
44 more elements
A system for summarizing a video spatially and temporally, comprising:
- means for specifying a spatial query for a video acquired of a scene by a camera, in which the spatial query specifies a spatially contiguous set of pixels with a particular frame of the video;
- means for specifying a temporal query for the video, in which the temporal query specifies a temporally contiguous set of pixels in multiple adjacent frames;
- means selecting frames of the a video according to the spatial query and the temporal query;
- and
- means compositing the selected frames into an image that represents spatial and temporal characteristics of the scene, in which the compositing is according to a predetermined order and predetermined opacities, in which the compositing is according to:
- ∑
- i
- =
- 1
- n
-
- α
- i
- ×
- i
- n
- =
- 1
- ,
-
- and
-
-
- α
- i
- ×
- i
- n
- =
- α
- j
- ×
- j
- n
-
-
- for
-
-
- all
-
-
- i
- ≠
- j
- ,
- where αi is an opacity for an ith layer in the image, and n represents a number of selected images.
44 more elements
Dependencies are inferred from the claim text.