Suppress pixel coloration errors in HDR video systems
The present disclosure is directed to a process to partially or fully suppress, or limit, pixel coloration errors. These pixel coloration errors, represented by small noise values, can be introduced during signal processing of high dynamic range (HDR) video signals. Converting visual content to a half precision floating point representation, for example, FP16, can introduce small amounts of signal noise due to value rounding. The noise can be multiplied and accumulated during HDR signal processing resulting in visual artifacts and degraded image quality. The disclosure can detect these noise amounts in pixel color component values, and suppress, or partially suppress, the noise to prevent the noise from accumulating during subsequent HDR signal processing.
1. A method to suppress pixel color component noise in a high dynamic range (HDR) signal, comprising:
quantizing a color value representation of a pixel to a color value representation having less precision, wherein said quantizing creates a noise level in one or more color component values of said pixel;
comparing said color component values to a noise threshold for said pixel; and
reducing, utilizing said comparing, one or more of said color component values of said pixel, wherein said noise threshold is computed from said noise level.
2. The method as recited in claim 1 , wherein said reducing clamps said color component value to zero, when said color component value is less than said noise threshold.
3. The method as recited in claim 1 , wherein said reducing occurs, during a processing of said HDR signal, after a color space conversion (CSC) process and prior to a perceptual quantizer (PQ) inverse electro-optical transfer function process.
4. The method as recited in claim 1 , wherein a spatial or temporal pixel processing is performed on said HDR signal.
5. The method as recited in claim 1 , wherein said noise threshold utilizes a noise factor and a maximum value from said color component values.
6. The method as recited in claim 1 , where said noise threshold utilizes a noise factor and a luminescence value computed from said color component values.
7. The method as recited in claim 6 , wherein said noise factor is 1/2048.
8. The method as recited in claim 7 , further comprising:
increasing said noise factor, where a noise accumulation exceeds a target value.
9. The method as recited in claim 1 , wherein said noise threshold is a negative noise value, and wherein said reducing one or more of said color component values of said pixel comprises multiplying each of said one or more said color component values of said pixel with said noise threshold.
10. The method as recited in claim 9 , wherein said reducing is combined with a scRGB to Rec2020RGB conversion process.
11. The method as recited in claim 9 , wherein said reducing clamps said color component value to zero when said color component value is less than an absolute value of said noise threshold.
12. A computer program product having a series of operating instructions stored on a non-transitory computer-readable medium that directs a data processing apparatus when executed thereby to perform operations to suppress noise in color component values of a pixel for a high dynamic range (HDR) signal, said operations comprising:
quantizing a color value representation of a pixel to a color value representation having less precision, wherein said quantizing creates a noise level in one or more color component values of said pixel;
comparing said color component values to a noise threshold for said pixel; and
reducing, utilizing said comparing, one or more of said color component values of said pixel, wherein said noise threshold is computed from said noise level.
13. The computer program product as recited in claim 12 , wherein said reducing clamps said color component value to zero, where said color component value is less than an absolute value of said noise threshold.
14. The computer program product as recited in claim 12 , wherein said reducing occurs, during a processing of said HDR signal, after a color space conversion (CSC) process and prior to a perceptual quantizer (PQ) inverse electro-optical transfer function process.
15. The computer program product as recited in claim 12 , wherein a spatial or temporal pixel processing is performed on said HDR signal.
16. The computer program product as recited in claim 12 , wherein said noise threshold utilizes a noise factor, and a maximum value from said color component values or a luminescence value computed from said color component values.
17. The computer program product as recited in claim 12 , wherein said noise threshold is a negative noise value, and wherein said reducing one or more of said color component values of said pixel comprises multiplying each of said one or more said color component values of said pixel with said noise threshold.
18. The computer program product as recited in claim 12 , further comprising:
increasing a noise factor used to calculate said noise threshold, where a noise accumulation exceeds a target value.
19. A noise suppression system for color component values of a pixel, comprising:
a receiver that receives a high dynamic range (HDR) content as an HDR signal; and
a signal processor that executes processes on said HDR signal to generate an adjusted HDR signal,
wherein one of said processes suppresses noise for one or more of said pixels of said HDR signal by comparing a calculated noise threshold to said color component values of said pixel to clamp to a value of zero, one or more of said color component values of said pixel, and
wherein said calculated noise threshold is computed utilizing a noise level of said pixel and a noise factor and said signal processor analyzes a noise accumulation and increases said noise factor to compensate for said noise accumulation.
20. The noise suppression system as recited in claim 19 , further comprising:
a transmitter that transmits said adjusted HDR signal.
21. The noise suppression system as recited in claim 19 , wherein said processes comprise at least one processor that performs spatial pixel processing or temporal pixel processing.
22. The noise suppression system as recited in claim 19 , wherein said signal processor performs matrix operations and combines one or more HDR signal processes.