Luminance and chromaticity measurement with 3D object maps

Abstract

Recent studies have proposed the method for measuring chromaticity and luminance distributions of a natural scene with a digital camera. In this method, using multiple images of the scene taken with different exposures, a high dynamic range image is created. By analyzing the created image, chromaticity and luminance distributions are acquired. By combining this method and Kinect that can acquire depth information of objects, the present study proposes the method for measuring the chromaticity and luminance distributions together with a 3D map of objects in a space. We conducted an experiment in which two images were taken by a digital camera and Kinect cameras. On the basis of point correspondences over the two images, a transformation matrix was computed. We examined the accuracy of the matching between the image obtained by Kinect and that transformed from a digital camera image.