Standardized spectral and radiometric calibration of consumer cameras

Abstract

Consumer cameras, particularly onboard smartphones and UAVs, are now commonly used as scientific instruments. However, their data processing pipelines are not optimized for quantitative radiometry and their calibration is more complex than that of scientific cameras. The lack of a standardized calibration methodology limits the interoperability between devices and, in the ever-changing market, ultimately the lifespan of projects using them. We present a standardized methodology and database (SPECTACLE) for spectral and radiometric calibrations of consumer cameras, including linearity, bias variations, read-out noise, dark current, ISO speed and gain, flat-field, and RGB spectral response. This includes golden standard ground-truth methods and do-it-yourself methods suitable for non-experts. Applying this methodology to seven popular cameras, we found high linearity in RAW but not JPEG data, inter-pixel gain variations >400% correlated with large-scale bias and read-out noise patterns, non-trivial ISO speed normalization functions, flat-field correction factors varying by up to 2.79 over the field of view, and both similarities and differences in spectral response. Moreover, these results differed wildly between camera models, highlighting the importance of standardization and a centralized database. accuracy of color measurements and their conversion to standard measures, such as the CIE 1931 XYZ and CIELAB color spaces, is limited by distortions in the observed colors [20] and differences in spectral response functions [21,[23][24][25].Extensive (spectro-)radiometric calibrations of consumer cameras are laborious and require specialized equipment and are thus not commonly performed [23,54,60]. A notable exception is the spectral and absolute radiometric calibration of a Raspberry Pi 3 V2 webcam by Pagnutti et al. [57], including calibrations of linearity, exposure stability, thermal and electronic noise, flat-field, and spectral response. Using this absolute radiometric calibration, digital values can be converted into SI units of radiance. However, the authors noted the need to characterize a large number of these cameras before the results could be applied in general. Moreover, certain calibrations are device-dependent and would need to be done separately on each device. Spectral and radiometric calibrations of seven cameras, including the Raspberry Pi, are given in [51]. These calibrations include dark current, flat-fielding, linearity, and spectral characterization. However, for the five digicams included in this work, JPEG data were used, severely limiting the quality and usefulness of these calibrations, as described above.Spectral characterizations are more commonly published since these are vital for quantitative color analysis. Using various methods, the spectral responses of several Canon [1,19,23,24,46,51,54,61], Nikon [1,13,23,24,54,[59][60][61][62], Olympus [23, 24, 51], Panasonic [46], SeaLife [13], Sigma [60], and Sony [1, 23, 46, 51, 61] digital cameras (digicams), as well as a number of smartpho...

A universal smartphone add-on for portable spectroscopy and polarimetry: iSPEX 2

“…26 Data are obtained in RAW image format because of its high linearity and dynamic range. 18 This is in contrast to the JPEG images taken with the original, where nonlinearity and white balance introduced significant problems. 17,18 Two examples, taken with a 3D printed prototype, are shown in Fig.…”

“…9b), which is 390-700 nm. 18 These foils are produced by stretching transparent sheets of polymer, such as polyvinyl alcohol (PVA), polycarbonate (PC) or poly(methyl methacrylate) (PMMA). 24 While the pre-fabricated foils provide the desired retardance with great consistency, they are prohibitively expensive for low-cost CS purposes.…”

“…Data are processed according to the SPECTACLE method, 18 which was originally developed for iSPEX 2. A Python library for data processing specific to iSPEX 2 is currently in development.…”

“…The RGBG 2 channels of the Bayer-filter camera are treated separately rather than combined through interpolation, since interpolated data add no extra information. 18 After demosaicing, there are eight separate spectra, namely the combinations of ±Q and RGBG 2 . Fig.…”

“…The wavelength-calibrated spectra are corrected for the RGBG 2 spectral response functions (SRFs), again using SPECTACLE. 18 The SRFs are interpolated to the same wavelengths as the data, after which the data are divided by the SRF. This gives radiances, a constant factor away from absolute radiometric units, 18 as shown in Fig.…”