Iterative algorithms have been investigated for reconstructing images from data acquired with a new type of gamma camera based upon an electronic method of collimating gamma radiation. The camera is composed of two detection systems which record a sequential interaction of the emitted gamma radiation. Coincident counting in accordance with Compton scattering kinematics leads to a localization of activity upon a multitude of conical surfaces throughout the object. A two-stage reconstruction procedure in which conical line projection images as seen by each position sensing element of the first detector are reconstructed in the first stage, and tomographic images are reconstructed in the second stage, has been developed. Computer simulation studies of both stages and first-stage reconstruction studies with preliminary experimental data are reported. Experimental data were obtained with one detection element of a prototype germanium detector. A microcomputer based circuit was developed to record coincident counts between the germanium detector and an uncollimated conventional scintillation camera. Point sources of Tc-99m and Cs-137 were used to perform preliminary measurements of sensitivity and point spread function characteristics of electronic collimation.

We introduce a technique to fill large holes in LiDAR data sets. We combine concepts from patch-based image inpainting and gradient-domain image editing to simultaneously fill both texture and structure in a LiDAR scan. We discuss the problems with directly inpainting a depth image, and present a solution to this problem based on inpainting the depth gradients. Once the inpainted depth gradients are obtained, we use an image reconstruction technique to obtain the final 3D scene structure. We present several realworld examples of this technique with excellent results.

No abstract

No abstract

No abstract