Imaging colorimetry using a digital camera with dental applications and digital video capture image analysis.

摘要

In this work, we have investigated three different research problems. In the first chapter, we introduce an imaging colorimeter for measuring tooth color that is based on two exposures with a digital camera using two different filters. We describe new procedures for alignment and calibration of the system. We also consider three different sources of illumination. We propose two approaches to estimating tooth color information from the acquired camera data, both of which employ linear regression. In the first approach, we estimate CIE XYZ coordinates of the tooth under a pre-selected illuminant. In the second approach, we estimate the spectral reflectance of the tooth, thus enabling us to predict tristimulus values under any illuminant. Finally, we present results from in vivo measurement of subject's teeth.;In the second chapter, a subspace matching method is developed to design color filters for a multi-spectral imaging system to capture the spectral reflectance of surfaces accurately with minimal number of measurements. The imaging colorimeter developed in chapter 1 is used as an example in the simulation where the performance of the best filters designed with our method and Wolski's [1] method are compared.;The third chapter describes the development of an automated prescreening tool, along with an image fidelity metric to reduce the workload of human observers involved in the image screening. This tool is capable of performing automatic image registration, configuration page identification and time stamp detection. A visualization module is also included to highlight regions of interest in an image. This tool works for a wide range of image types and content.