Can we use smartphone-imaging sensors as low cost luminance mapping tools to support design processes, integrated lighting system control, and human factors research?

摘要

In this paper, High Dynamic Range (HDR) Photography methods (Inanici, 2006) are employed with both smartphone cameras and full-frame Complementary Metal-Oxide Semiconductor (CMOS) sensors to evaluate their accuracy in creating luminance maps and usefulness in practical applications. We answered three questions: can smartphone quality cameras function as inexpensive imaging sensors for HDR applications to collect scene luminance information accurate enough for use in: 1) human factors research, 2) building automation and control systems, and 3) teaching and learning scenarios? Three popular smartphones, the iPhone 5s, iPhone 6s, and Galaxy S6 Edge+ were chosen for comparison to the full-frame CMOS sensor found in the Canon EOS 5D, following current HDR best practices. In order to study each camera's absolute and relative accuracy across a wide brightness range we collected data within high, medium, and low luminance source scenes. We compared HDR luminance results captured by the iPhone 5s, iPhone 6s, and Galaxy S6 Edge+ in each scene with HDR luminance results captured using the Canon EOS 5D camera with a Sigma 8mm F3.5 EX DG fisheye lens, and also evaluated both against hand held measurements recorded using Minolta LS-110 and LS-160 luminance meters. When comparing luminance data collected with the smartphones to data collected using the Canon EOS 5D camera we found that the smaller CMOS sensors used in smartphones are more sensitive and tend to have more problems with luminance overflow (Jakubiec et al., 2016). Our findings suggest that smartphone cameras presently have severely limited value for use in human factors visual comfort research, especially when measuring the solar corona or sunlit surfaces, and for similar reasons should not be used to control automated blinds. The maximum luminance in a high brightness scene captured by the iPhone 6s is 13,790 cd/m~2 while the Galaxy S6 Edge+ captured a maximum of 51,580 cd/m~2, effectively capturing a mere 1.7% and 6.2% of the Canon EOS 5D's 825,300 cd/m~2 maximum respectively. However, these inexpensive cameras may hold potential for use in the control of electric lighting due to relatively lower luminance value ranges needed to adjust electric lighting. Smartphone cameras could also be used in basic teaching applications where the relative brightness of surfaces rather than absolute brightness is of interest.