Aptina is committed to sharing its world class expertise in system image quality evaluation with the imaging community so that better imaging products can be made to meet the needs of the end users.
Digital reference stimuli and softcopy quality ruler software (February 2012)
The Digital Reference Stimuli (DRS) are a set of images that when suitably displayed and viewed, have known absolute quality, expressed in the Standard Quality Scale (SQS) of ISO 20462-3. The DRS comprise 31 images each of 21 different scenes, which vary in modulation transfer function and thus sharpness and quality, with increments between adjacent images averaging about one just noticeable difference (JND) of quality. The images are primarily intended for use in the softcopy quality ruler defined in ISO 20462-3. The softcopy ruler enables efficient calibrated measurement over a wide range of image quality, using a graphical user interface with a slider for selecting a ruler image with quality equal to that of a test image. Freeware Matlab® code for running softcopy ruler experiments is provided in addition to extensive supporting documentation. The DRS and softcopy ruler code were developed during the International Imaging Industry Association’s Camera Phone Image Quality initiative, and ISO 20462 is currently being revised to reflect these advances.
Aptina noise study archived data (December 2010)
A system simulation model was used to create scene-dependent noise masks that reflect current performance of mobile phone cameras. Stimuli with different overall magnitudes of noise and with varying mixtures of red, green, blue, and luminance noises were included in the study. Eleven treatments in each of ten pictorial scenes were evaluated by twenty observers using the softcopy ruler method. In addition to determining the quality loss function in just noticeable differences (JNDs) for the average observer and scene, transformations for different combinations of observer sensitivity and scene susceptibility were derived. Test targets in linear sRGB and rendered L*a*b* spaces for each treatment are made available here to enable other researchers to test metrics of their own design and calibrate them to JNDs of quality loss without performing additional observer experiments. Such JND-calibrated noise metrics are particularly valuable for comparing the impact of noise and other attributes, and for computing overall image quality.