Speakers
Description
Low contrast images are not properly viewable with reflective panels under bright ambient light condition. Image histogram (pixel value vs count plot) can be used for estimating contrast. High contrast images have histogram spread across pixel values. Various equalization techniques are used for spreading histogram of low contrast images over the pixel values. This generated Histogram is notifiedto the user via histogram udev event.
Mutter upon receiving the histogram event goes on to read the histogram data from KMD via exposed crtc property.
A library is been developed to take this generated histogram as an input and apply a logical algorithm to generate an Image EnhancemenT(IET) Factor. This is further fed back to the KMD via crtc property. KMD will feed this back to the hardware. Hardware will use this as a multiplicand factor to multiply with the incoming pixels at the end of the pipe which is then pushed onto the display.
This library "Global Histogram Enhancement" will take the histogram and other parameters as input and apply the algorithm to enhance the pixel density and returns the enhanced factor.
| Category | Design of Users and Developer Experiences |
|---|---|
| Talk Description | Low contrast images are not properly viewable with reflective panels under bright ambient light condition. Image histogram (pixel value vs count plot) can be used for estimating contrast. High contrast images have histogram spread across pixel values. Various equalization techniques are used for spreading histogram of low contrast images over the pixel values. This generated Histogram is notifiedto the user via histogram udev event. Mutter upon receiving the histogram event goes on to read the histogram data from KMD via exposed crtc property. A library is been developed to take this generated histogram as an input and apply a logical algorithm to generate an Image EnhancemenT(IET) Factor. This is further fed back to the KMD via crtc property. KMD will feed this back to the hardware. Hardware will use this as a multiplicand factor to multiply with the incoming pixels at the end of the pipe which is then pushed onto the display. This library "Global Histogram Enhancement" will take the histogram and other parameters as input and apply the algorithm to enhance the pixel density and returns the enhanced factor.This Histogram enhancement can be furthermore advanced to save power, as Display power saving technology DPST is composed of three blocks : Image Statistics - A block in HW that generates image statistics based on the pixel stream input. These stats are used by Processing block to determine how much the backlight can be lowered. Pixel Enhancement - A block in HW that increases the pixel values to compensate for the brightness loss due to lowering of the backlight level. Processing - This is a SW block that determines how much the backlight level can be reduced, then sets the backlight level and the Pixel Enhancement block |
| Author(s) Bio | GPU Software development engineers at Intel |
| Pronouns | he |
| Where are you located? | Bangalore |
| Do you need travel sponsorship from GNOME Foundation in order to join our event? | No |
