Orthogonal Tensor Neighborhood Preserving Embedding (OTNPE) is an efficient dimensionality reduction algorithm for two-dimensional images. However, insufficiencies of the robustness performance and deficiencies of supervised discriminant information are remained. Motivated by the sparse learning, an algorithm called Orthogonal Tensor Sparse Neighborhood Embedding (OTSNPE) for two-dimensional im...

Kamasak, Mustafa E. Ph.D., Purdue University, May, 2005. Statistical Methods for Reconstruction of Parametric Images from Dynamic PET Data. Major Professor: Charles A. Bouman. In this thesis, we demonstrate and evaluate a new and efficient method for direct voxel-wise reconstruction of kinetic parameter images using all frames of the PET data. The direct parametric image reconstruction is formu...

This doctoral thesis proposes a new algorithm for the reconstruction of impedance images in monitored objects. The algorithm eliminates the spatial resolution problems present in existing reconstruction methods, and, with respect to the monitored objects, it exploits both the partial knowledge of configuration and the material composition. The discussed novel method is designed to recognize cer...

An SPECT image can be approximated as the convolution of the ground truth spatial radioactivity with the system point spread function (PSF). The PSF of an SPECT system is determined by the combined effect of several factors, including the gamma camera PSF, scattering, attenuation, and collimator response. It is hard to determine the SPECT system PSF analytically, although it may be measured exp...

We present a novel multi-view 3D reconstruction algorithm which unifies the advantages of several recent reconstruction approaches. Based on a known environment causing occlusions and on the cameras' pixel grid discretization, an irregular partitioning of the reconstruction space is chosen. Reconstruction artifacts are rejected by using plausibility checks based on additional information about ...

Astronomical optical interferometers sample the Fourier transform of the intensity distribution of a source at the observation wavelength. Because of rapid perturbations caused by atmospheric turbulence, the phases of the complex Fourier samples (visibilities) cannot be directly exploited. Consequently, specific image reconstruction methods have been devised in the last few decades. Modern poly...

Recclltly we proposed a 3D reconstruction algorithm based upon Fourier analysis using Cartesian coordinates. In this algorithm thc computations to determine the values of the 3D Discrete Fourier Transforms of the density of an asymmetric object could be naturally distributed over the nodes of a parallel system. Now we present an improvement of this algorithm which, for reconstruction at poinls ...

This paper proposes a learning scheme based still image super-resolution reconstruction algorithm. Superresolution reconstruction is proposed as a binary classification problem and can be solved by conditional class probability estimation. Assuming the probability takes the form of additive logistic regression function, AdaBoost algorithm is used to predict the probability. Experiments on face ...

Random equivalent sampling (RES) can composite a waveform with high equivalent sampling rate from multiple low speed sampling sequences. In practical application, the performance of RES signal reconstruction would be degraded by the non-uniform distribution of sampling time. Compressed sensing (CS) theory is adopted to reconstruct RES samples, which could mitigate the inherent coherence of samp...