Passive color night vision
AssigneeFyotech LLC
InventorHans H. Uhlig
Life of this patent
Estimating the term…
Abstract
A novel Passive Color Night Vision (PCNV) that provides a fully integrated dynamically controlled passive color night vision algorithm to work in both daylight or in low light conditions. The PCNV can dynamically switch or adjust between lighted vision and night vision. The PCNV is a passive solution that does not require active illumination by the use of an IR emitter diode as found in the prior art night vision solutions. Visible and IR spectrum composite image data manipulation is performed to achieve true color in low light conditions.
Claims16
A method for colorization of a digital image captured in low light conditions, the method comprising the actions of:
- capturing raw image data of an image from an image sensor, the data including RGB values and IR values;
- converting the raw image data to a digital RGB file containing a viewable color image by utilizing the RGB values and the IR values;
- normalizing the RGB values by augmenting the RGB values with the IR value multiplied by a luma matrix to bias the values towards the middle of the visible spectrum;
- augmenting the normalized RGB values by applying and adjustment for each RGB value with a 3D lookup table function;
- and
- outputting a final colorized image.
A system configured to colorize a digital image captured in low light conditions, the system comprising the actions of:
- an image sensor configured to capture raw image data of an image, the raw image data including RGB values and IR values;
- a conversion algorithm configured to convert converting the raw image data to a digital RGB file containing a viewable color image by utilizing the RGB values and the IR values;
- a normalizing algorithm configured to normalize the RGB values by augmenting the RGB values with the IR value multiplied by a luma matrix to bias the values towards the middle of the visible spectrum;
- and
- a 3D lookup table function configured to augment the normalized RGB values by applying and adjustment for each RGB value with a 3D lookup table function.
Dependencies are inferred from the claim text.