Study Of Luminous Environment Measurement Through High Dynamic Range Imaging

The traditional method of luminous environment measurement uses the luminance meter scanning a scene point by point to obtain the illumination distribution.That method is time-consuming and with high operation requirements,which may introduce errors due to the instability of light source or the change of measurement position.Therefore,a quick and convenient method is necessary to measure the luminous environment of a scene.With the development of imaging and computer technology,the application of cameras for imaging measurement has attracted more and more attention.And the high information storage capacity of high dynamic range images enables users to capture the various luminous parts of a scene simultaneously,making the quick luminous environment measurement possible.However,the commonly used high dynamic range imaging method is flawed for the accurate expression of photometric and colorimetric information,and the variation of camera parameters needs to be calibrated individually for different images,which affects the application and popularization of the technology.In this study,we focused on the problem of photometric and colorimetric expression in high dynamic range imaging and the complicated operation of actual scene measurement,and put forward an imaging approach of luminous environment measurement.Combined with color science and high dynamic range imaging technology,the study applied a series of methods such as noise analysis,lens correction,camera spectral sensitivity recovery,camera parameters independent colorimetric characterization to enhance the accurate expression of photometric and colorimetric information.Finally,an application instance was implemented to illustrate the achievements of this study.The study took a start from analyzing the sensor noise and lens,non-uniformity so as to improve the performance of each part in the imaging progress.Then,the camera's integrated response function was deduced by applying the spectral recovery algorithm to traditional camera model added with nonlinear parameter and color channel crosstalk matrix.And this function is a connection between the camera response and the spectral radiance of the scene luminance and so can be used for further photometric and colorimetric application.Derived from the camera response function and root polynomial color correction method,a different colorimetric characterization method of digital cameras was proposed,requiring no consideration of the camera parameters.The method was aimed at improving the accuracy of color expression of high dynamic range imaging,and coping with the changes of camera parameters in the measuring process,which adopted unit response values,i.e.radiance flux surface density of color channels,during the procedure.And the feasibility of this method was verified by the colorimetric characterization experiments with the low dynamic range images of different camera parameters and two sets of high dynamic range images.In order to further demonstrate the practicality of the achieved outcomes in this study,an instance of indoor luminous environment measurement was presented.The color correction of high dynamic range images based on illumination response values was performed by employing the camera response function and the stored color checker's SPD data to estimate the colorimetric transformed matrix for high dynamic range imaging.And then the distribution of luminance as well as colorimetric coordinates was predicted for the actual scene through the high dynamic range image's colorimetric parameters.Finally,the main contents of this dissertation were briefly summarized,and the perspectives of the future work were also forecasted.