2015-Acutal: Optical Engineering – Aspöck Portugal
2013-2014: ERASUMS+ at Adam Mickiewicz University, Poznan (Poland)
2010-2016: Engineering Physics – Integrated Master
Photorealistic Luminance Measurement Method and Homogeneity Analysis by Using Digital Cameras
Luminance homogeneity of automotive rear signal functions started some years ago on tail but is now emerging as mandatory on other rear functions such as turn indicators and stops. Some mathematical methods to define homogeneity coefficients or indexes have been developed, though there is a lack of quick analysis technology to evaluate luminance maps.
This paper presents a photorealistic luminance measurement method and homogeneity analysis by using a digital camera as a luminance meter. Tests were performed in an experimental setup by using a reflective diffuse white sample, a red, white and yellow automotive LEDs, a calibrated luminance meter (LMT L-1003) and a digital camera (Olympus E-PL2). Luminance measurement was acquired by varying the input power of the LEDs and results were interpolated to find the OECF (Opto-Electronic Conversion Function) of the digital camera sensor. This calibration process was also optimized for saturated colours, such as red environments, where the sensor in most sensible and most of the conversion methods fail.
Then, image processing routines were developed in the ImageJ software for accurate conversion of 24-bit RGB images into luminance values. At that point, two homogeneity class analysis were defined, one that operates over the entire image or photography and other that operates over a region of interest only. Both classes offer pixel by pixel mapping or neighbour pixel averaging, depending on part luminance distribution and complexity. These two classes were sub-divided into two calculation methods, leading to a very quick and precise method of luminance homogeneity characterisation.
Feedback from users that included this routine on the development workflow, indicate that not only reduce the number of optimization loops, but also creates quantitative objectives for part homogeneity. This is also a useful tool to compare renderings, light mock-ups and production parts lit appearance.