Quality assessment for images that have extended dynamic ranges or wide color gamuts

摘要

A first video signal is accessed, and represented in a first color space with a first color gamut, related to a first dynamic range. A second video signal is accessed, and represented in a second color space of a second color gamut, related to a second dynamic range. The first accessed video signal is converted to a video signal represented in the second color space. At least two color-related components of the converted video signal are mapped over the second dynamic range. The mapped video signal and the second accessed video signal are processed. Based on the processing, a difference is measured between the processed first and second video signals. A visual quality characteristic relates to a magnitude of the measured difference between the processed first and second video signals. The visual quality characteristic is assessed based, at least in part, on the measurement of the difference.

主权项

1. A method, comprising the steps of:
accessing a first video signal that is represented in a first color space with a first color gamut which is related to a first dynamic range, and a second video signal that is represented in a second color space of a second color gamut, wherein the second color space is associated with a second dynamic range; converting the accessed first video signal to a converted video signal that is represented in the second color space of the second color gamut with the second dynamic range; processing the converted video signal and the accessed second video signal, wherein for at least one frame of the converted video signal and a corresponding frame of the second video signal the processing step comprises:
applying a non-linear scaling function to the converted video signal to generate a first scaled video signal;applying the non-linear scaling function to the second video signal to generate a second scaled video signal; andapplying a photographic mapping function to the first scaled video signal and the second scaled video signal to generate a first mapped video signal and a second mapped video signal; and based at least in part on the processing step, measuring a difference between the processed first mapped and second mapped video signals, wherein measuring the difference comprises:
generating based at least in part on the first mapped video signal and the second mapped video signal a luminance comparison metric, a contrast comparison metric, and a structural comparison metric; andcombining the luminance comparison metric, the contrast comparison metric, and the structural comparison metric to generate a quality metric which indicates how pixels of the first video signal match corresponding pixels of the second video signal.