研究了一种基于高动态范围(HDR)图像的亮度测量方法。基于数码相机的感光特性及成像原理,采用Photosphere软件将一组低动态范围时间序列图像合成为HDR图像,获取该图像的灰度值并与经线性变换得到的3个刺激值RGB数据进行拟合,得到亮度的计算模型,并对3种背景亮度下标准色板各色块亮度进行计算。研究结果表明,3种背景亮度下计算得到的标准色板各色块的亮度平均误差在9.22%以内;随着背景亮度的降低,所有色块目标和彩色目标的亮度误差均减小,灰度目标的亮度误差在背景亮度适中时达到最小。

Based on the high dynamic range (HDR) image technology, a calculation method for obtaining the luminance parameters is investigated. In this method, according to the photographic characteristics and the imaging principle of digital cameras, a set of low dynamic range time series images is synthesized into a HDR image using the Photosphere software. The gray value of this image is calculated and fitted with the tristimulus value of R, G, B data obtained by linear transformation. And thus the luminance is obtained. The luminance of each color block in the standard color plate is calculated under three background luminance and it is found that the average luminance error is less than 9.22%. Moreover, the luminance errors of all color block targets and coloured targets are reduced with the decrease of background luminance, and the luminance error of the gray targets reaches a minimum when the background luminance is within an appropriate scope.