Back to results
US 2015/0189304 A1Published application
Multi-spectral image compression
AssigneeGoogle Technology Holdings LLC
InventorsJohn T. Bretscher, Randall S. Vaas
Life of this patent
Estimating the term…
Abstract
This document describes techniques and apparatuses for implementing multi-spectral compression. These techniques enable a computing device to capture a set of images at different spectral sensitivities using a multi-spectral array imager. Once the set of images are captured by the multi-spectral array imager, the set may be encoded using encoding techniques that involve spectral prediction in addition to spatial and temporal prediction.
Claims20
A computer-implemented method comprising:
- obtaining a set of images of a scene that were captured substantially simultaneously by an array imager, the set of images including an image that is captured with a different spectral sensitivity than an other image;
- encoding information representative of the image using spatial prediction;
- and
- encoding other information representative of the other image using spectral prediction, the spectral prediction predicting the other information representative of the other image based, in part, on the information representative of the image and a model representing the difference in spectral sensitivities between the image and the other image.
1 more element
A device comprising:
- a multi-spectral array imager configured to capture a set of images of a scene substantially simultaneously, the multi-spectral array imager further configured to capture at least two of the images in the set with different spectral sensitivities;
- and
- a compression engine embodied at least partially in hardware of the device and configured to:
- cause information representative of a first of the at least two images to be encoded using spectral prediction, the spectral prediction predicting the information representative of the first image based, at least in part, on other information representative of a second of the at least two images that is captured with a different spectral sensitivity than the first image.
1 more element
A computer-implemented method comprising:
- obtaining encoded information representative of an image from a set of images of a scene, the image captured with a different spectral sensitivity than another image from the set due to the image and the other image being captured using different spectral filters;
- and
- decoding the encoded information based on a determination that spectral prediction was used to encode the information representative of the image, the spectral prediction predicting the information representative of the image based, in part, on information representative of the other image from the set and a model representing the difference in spectral sensitivities between the image and the other image.
Dependencies are inferred from the claim text.