Combined three-dimensional computer vision and epi-illumination fluorescence imaging system

Most of the reported fluorescence imaging methods and systems highlight the need for three-dimensional information of the inspected region surface geometry. The scope of this manuscript is to introduce an epi-illumination fluorescence imaging system, which has been enhanced with a binocular machine vision system for the translation of the inverse problem solution to the global coordinates system. The epi-illumination fluorescence imaging system is consisted of a structured scanning excitation source, which increases the spatial differentiation of the measured data, and a telecentric lens, which increases the angular differentiation. On the other hand, the binocular system is based on the projection of a structured light pattern on the inspected area, for the solution of the correspondence problem between the stereo pair. The functionality of the system has been evaluated on tissue phantoms and calibration objects. The reconstruction accuracy of the fluorophores distribution, as resulted from the root mean square error between the actual distribution and the outcome of the forward solver, was more than 80%. On the other hand, the surface three-dimensional reconstruction of the inspected region presented 0.067±0.004 mm accuracy, as resulted from the mean Euclidean distance between the three-dimensional position of the real world points and those reconstructed.