洋红
放大倍数
RGB颜色模型
人工智能
光学
计算机视觉
渲染(计算机图形)
色度
显微镜
数学
计算机科学
材料科学
物理
墨水池
语音识别
作者
Simen Alexander Linge Johnsen,Jörg Bollmann,H.W. Lee,Y. ZHOU
摘要
Summary Here a work flow towards an accurate representation of interference colours (Michel‐Lévy chart) digitally captured on a polarised light microscope using dry and oil immersion objectives is presented. The work flow includes accurate rendering of interference colours considering the colour temperature of the light source of the microscope and chromatic adaptation to white points of RGB colour spaces as well as the colour correction of the camera using readily available colour targets. The quality of different colour correction profiles was tested independently on an IT8.7/1 target. The best performing profile was using the XYZ cLUT algorithm and it revealed a ΔE00 of 1.9 (6.4 no profile) at 5× and 1.1 (8.4 no profile) at 100× magnification, respectively. The overall performance of the workflow was tested by comparing rendered interference colours with colour‐corrected images of a quartz wedge captured over a retardation range from 80–2500 nm at 5× magnification. Uncorrected images of the quartz wedge in sRGB colour space revealed a mean ΔE00 of 12.3, which could be reduced to a mean of 4.9 by applying a camera correction profile based on an IT8.7/1 target and the Matrix only algorithm (ΔE00 < 1.0 signifies colour differences imperceptible by the human eye). ΔE00 varied significantly over the retardation range of 80–2500 nm of the quartz wedge, but the reasons for this variation is not well understood and the quality of colour correction might be further improved in future by using custom made colour targets specifically designed for the analysis of high‐order interference colours.
科研通智能强力驱动
Strongly Powered by AbleSci AI