Quantitative measurement of contrast, texture, color, and noise for digital photography of high dynamic range scenes

Abstract

Today, most advanced mobile phone cameras integrate multi-image technologies such as high dynamic range (HDR) imaging. The objective of HDR imaging is to overcome some of the limitations imposed by the sensor physics, which limit the performance of small camera sensors used in mobile phones compared to larger sensors used in digital single-lens reflex (DSLR) cameras. In this context, it becomes more and more important to establish new image quality measurement protocols and test scenes that can differentiate the image quality performance of these devices. In this work, we describe image quality measurements for HDR scenes covering local contrast preservation, texture preservation, color consistency, and noise stability. By monitoring these four attributes in both the bright and dark parts of the image, over different dynamic ranges, we benchmarked four leading smartphone cameras using different technologies and contrasted the results with subjective evaluations.

Objective image quality evaluation of HDR videos captured by smartphones

“…The second setup, named HDR Composite (figure 1.b), was initially introduced in [1] and [2] to measure a device's capacity to capture and render static HDR scenes into SDR images. The setup contains two back-lit panels, each covered by a transmissive target that contains a color chart, a texture chart, and a gray scale of 63 uniform patches spanning a transmittance range of approximately 8EV.…”

“…reliance on measurement spaces that either have known limitations or have not been validated in the case of HDR videos, such as CIE-L*a*b*) ([1], [2] and [3]). • They have only been tested on still images content, and we lack data on their behavior when confronted with a dynamic scene ([1], [2]).…”

“…reliance on measurement spaces that either have known limitations or have not been validated in the case of HDR videos, such as CIE-L*a*b*) ([1], [2] and [3]). • They have only been tested on still images content, and we lack data on their behavior when confronted with a dynamic scene ([1], [2]). Additionally, a lot of previous works available in the literature concern perceptual evaluation of whether an existing HDR video is reproduced correctly on the target display (after potentially lossy encoding, transmission or decoding stages), but do not evaluate whether the capture of a scene into that HDR content was done in an optimal manner in the first place.…”

Automatic Noise Analysis on Still Life Chart

“…• The formula standardized by CPIQ [1] (V N CPIQ ) • The formula in discussion in ISO15739 and lastly proposed [4] (V N ISO ) • The formula used by DXOMARK [14] (V N DXOMARK )…”