John Chiverton

Automatic Bootstrapping and Tracking of Object Contours

A new fully automatic object tracking and segmentation framework is proposed. The framework consists of a motion-based bootstrapping algorithm concurrent to a shape-based active contour. The shape-based active contour uses finite shape memory that is automatically and continuously built from both the bootstrap process and the active-contour object tracker. A scheme is proposed to ensure that the finite shape memory is continuously updated but forgets unnecessary information. Two new ways of automatically extracting shape information from image data given a region of interest are also proposed. Results demonstrate that the bootstrapping stage provides important motion and shape information to the object tracker. This information is found to be essential for good (fully automatic) initialization of the active contour. Further results also demonstrate convergence properties of the content of the finite shape memory and similar object tracking performance in comparison with an object tracker with unlimited shape memory. Tests with an active contour using a fixed-shape prior also demonstrate superior performance for the proposed bootstrapped finite-shape-memory framework and similar performance when compared with a recently proposed active contour that uses an alternative online learning model.

Volumetric partial volume quantification via a statistical model of 3-D voxel gradient magnitude

3-D volumetric data sets suffer from partial volume (PV) effects due to the finite bandwidth of the digital sampling process. A variety of approaches have been developed to quantify the PV effect in PET, SPECT, NMR and CT imaging modalities. Amongst these, voxel gradient magnitude information, modeled as a Rician distribution, has been suggested as a useful adjunct for statistical PV correction in 2-1) data. However, many biomedical image acquisition processes provide contiguous image slices arising from an acquisition process, which can be approximated to be 3-D in terms of the digital sampling process. Thus, classifiers using models that utilize extra information from the transverse or third perpendicular direction, in this case 3-D gradient magnitude information, should possess superior performance over algorithms that utilize lower dimensional information (e.g. intensity or 2D gradient features). Therefore, analytically derived probability distributions are presented to describe the 3-D gradient magnitude for 3-D isotropic and anisotropic data sets. A Bayesian classification framework, utilizing the 3-D isotropic and anisotropic gradient magnitude expressions, is compared with other models, illustrating superior performance for 3-D volumetric data

Adaptive partial volume classification of MRI data

Tomographic biomedical images are commonly affected by an imaging artefact known as the partial volume (PV) effect. The PV effect produces voxels composed of a mixture of tissues in anatomical magnetic resonance imaging (MRI) data resulting in a continuity of these tissue classes. Anatomical MRI data typically consist of a number of contiguous regions of tissues or even contiguous regions of PV voxels. Furthermore discontinuities exist between the boundaries of these contiguous image regions. The work presented here probabilistically models the PV effect using spatial regularization in the form of continuous Markov random fields (MRFs) to classify anatomical MRI brain data, simulated and real. A unique approach is used to adaptively control the amount of spatial regularization imposed by the MRF. Spatially derived image gradient magnitude is used to identify the discontinuities between image regions of contiguous tissue voxels and PV voxels, imposing variable amounts of regularization determined by simulation. Markov chain Monte Carlo (MCMC) is used to simulate the posterior distribution of the probabilistic image model. Promising quantitative results are presented for PV classification of simulated and real MRI data of the human brain.

Quantifying the Partial Volume Effect in PET Using Benford's Law

Partial volume (PV) correction techniques in PET or SPECT represents a key step in image quantification methods. The PV effect arises because of the blurring induced by the imaging systems point spread function (PSF), producing intra-voxel mixing of the signals arising from different functional tissue classes. Quantification of this effect is often required to recover the mixing components within a group of voxels, from whence the true tissue concentration in a given volume or region can be estimated. In this work we consider a probabilistic methodology that uses a phenomenological distribution known as Benfords law to quantify the partial volume effect. We establish for the first time, that the probability distribution of voxels subjected to the PV effect in discrete volumetric data can be well described by Benfords law. The probabilistic framework devised here can be applied generically across different imaging modalities including PET and SPECT. Results from simulated data are presented, along with a PET phantom study utilizing registered processed CT data as ground truth, to determine the quality of the resulting probabilistic voxel classification scheme. For a water filled hot insert using a 5:1 insert:background activity concentration, we find an overall voxel RMS error of 3% (compared to ground truth) in the estimated voxel mixing vectors. This error rises to 8% for a cold air-filled insert in a warm background.

On-line Learning of Shape Information for Object Segmentation and Tracking

We present segmentation and tracking of deformable objects using non-linear on-line learning of high-level shape information in the form of a level set function. The emphasis is for successful tracking of objects that undergo smooth arbitrary deformations, but without the a priori learning of shape constraints. The high-level shape information is learnt on-line by defining a memory of object samples in a high-dimensional shape space. These shape samples are then used as weights via a locally defined shape space kernel function to define a template against which potential future shapes of the tracked object can be compared. Results for the successful tracking of a range of deformable motions are presented.

Mixture effects in FIR low-pass filtered signals

Accurate classification of signals composed of two or more classification classes (e.g., biomedical imaging data with pathological structures) might utilize a density that takes account of the signal acquisition process. A new density based on a Gaussian point spread function (PSF) and another utilizing a phenomenological observation known as Benfords Law are presented. Histograms of filtered signals are compared with these densities. The results suggest that the Gaussian PSF-based density is somewhat better than the Benfords Law density. Both approaches provide improved fits to histograms from data convolved with a variety of different PSFs over an existing mixture formulation.

Multiscale Shannon’s Entropy Modeling of Orientation and Distance in Steel Fiber Micro-Tomography Data

This paper is concerned with the modeling and analysis of the orientation and distance between steel fibers in X-ray micro-tomography data. The advantage of combining both orientation and separation in a model is that it helps provide a detailed understanding of how the steel fibers are arranged, which is easy to compare. The developed models are designed to summarize the randomness of the orientation distribution of the steel fibers both locally and across an entire volume based on multiscale entropy. Theoretical modeling, simulation, and application to real imaging data are shown here. The theoretical modeling of multiscale entropy for orientation includes a proof showing the final form of the multiscale taken over a linear range of scales. A series of image processing operations are also included to overcome interslice connectivity issues to help derive the statistical descriptions of the orientation distributions of the steel fibers. The results demonstrate that multiscale entropy provides unique insights into both simulated and real imaging data of steel fiber reinforced concrete.

Tracking with active contours using dynamically updated shape information

An active contour based tracking framework is described that generates and integrates dynamic shape information without having to learn a priori shape constraints. This dynamic shape information is combined with dynamic photometric foreground model matching and background mismatching. Boundary based optical flow is also used to estimate the location of the object in each new frame, incorporating Procrustes shape alignment. Promising results under complex deformations of shape, varied levels of noise, and closeto-complete occlusion in complex textured backgrounds are presented.

Effects of steel fibre-aggregate interaction on mechanical behaviour of steel fibre reinforced concrete

ABSTRACT This work investigated the effects of fibre type, dosage and maximum aggregate size on the mechanical behaviour of concrete reinforced with steel fibres. Hooked-end steel fibres with 50 and 60 mm length and aspect ratios (length/diameter) of 45, 65 and 80 were used with maximum sizes of coarse aggregate of 10 and 20 mm. The same mix proportions of concrete were used throughout the investigation. Flexural testing of 600 mm square panels was performed. Subsequently, cores were taken from these panels and X-ray computed tomography was used to analyse the positioning of fibres in hardened concrete. The experimental results show that the performance of steel fibre-reinforced concrete improved drastically when compared to plain concrete without fibres. Longer, thinner fibres and smaller aggregates were noted to give the best results.

Activity Recognition from Video Data Using Spatial and Temporal Features

A method to monitor elderly people in an indoor environment using conventional cameras is presented. The method can be used to identify peoples activities and initiate suitable actions as needed. The originality of our approach is in combining spatial and temporal contexts with the position and orientation for the detected person. Preliminary evaluation, based only on the first two features (spatial and temporal), achieved the accuracy over 60% in a realistic residential environment. Although the results are based on using only two out of the four proposed input features, they already demonstrate a promising improvement over using a single feature in isolation.