Research On Obtaining Light Parameters Based On The High Dynamic Range Image

The study is a basic research of the national natural science foundation project (No.51178299). Through reviewing the existing test methods of the lighting environmenton roads, it is founded that: the traditional test method has lots of disadvantages, suchas heavy workload, time-consuming, high-risk and so on. And the Low DynamicRange image method also has limitations on obtaining the Photometric Parameterswhen the Brightness Contrast of the environment exceeds a certain extent. Therefore,the paper presents the High Dynamic Range Image technique to solve the problems.Through reviewing the the high dynamic range image technique and its application inthe field of optics, combined with the features of the road light environment, themultiple exposure method was selected to obtain a high dynamic range image andRadiance formats as image storage format.Researched on the photosensitive characteristics and imaging of digital cameras, andprinciples of optics and colorimetry, the theories and methods of obtaining the lightenvironment parameters from high dynamic range images were established, and sowere a correspondingly software. Both the method and the traditional method wereused to test light environment. By SPSS analysis and error analysis of exprimentaldata, precision studies of the established method were carried out.The experimental studies on accuracy showed: error rate was not greater than10%inthe luminance information extraction of each measuring point and even better onaverage results(less than2%). For measuring the road surface illumination, with areliable reflection coefficient of pavement material, the error rate did not exceed10%;with brightness coefficient method, the average illuminance of the road surface wasbetter(less than2%). Calculating the illuminance of the light source formed at thehuman eye, the method was more reliable than directly measured by illuminometer.Finally, the method was used respectively on measuring the ordinary road andoverpass road in Tianjin. The practices showed that the method was feasible andefficient. Compared with other methods, the method is easy to operate, shorttime-consuming, and can record the complete lighting information.