In this paper we present a multiple particle filter framework for accurate real road tracking application, which fuses image features, GPS and map data. We represent the road as a set of connected arcs extracted from a digital map. The state space for the tracker not only contains global variables such as GPS coordinates, but also local variables such as road are parameters. Moreover, the dimension of the state space varies with the number of road arcs. A multiple particle filter framework is developed in order to: 1) improve sampling efficiency for a large state space; 2) cope with the variable dimension of the state space; 3) integrate image-based road tracking (local) and GPS (global) data. Each tracker applies the condensation filtering algorithm. We use multiple trackers for estimating state variables related to global positioning, camera and the lane, and road arcs respectively. Each tracker samples particles based on the state estimates by its predecessor. Experiments with real road videos demonstrate the effectiveness of the approach and the improvement to global positioning.
A good shape descriptor is necessary for automatically identifying landmarks on boundaries. Our method of boundary shape description is based on the notion of c- scale, which is a new local scale concept, defined at each boundary element. From this representation we can extract special points of interest such as convex and concave corners, straight lines, circular segments, and inflection points. The results show that this method gives a complete description of shape and allows the automatic positioning of mathematical landmarks, which agree with our intuitive ideas of where landmarks may be defined. This method is applicable to spaces of any dimensionality, although we have focused in this paper on 2D shapes.
Angiogenesis, the formation of de novo blood vessels, has been implicated in a long list of human disorders, including cancer, diabetes, and in recent years, neurodegenerative diseases, a list that continues to grow.Study of the formation of microvasculature, therefore, has important implications in diagnosis and treatment of diseases.This paper describes how we combine a CPUbased Cellular Potts model of sprouting angiogenesis in three dimensions with medical imaging techniques and graphics processing unit (GPU) accelerated fluid dynamics equations to create an individual-based angiogenesis simulation.The use of GPU, optimized for fast, highly parallel mathematical operations, provides an increase in simulation speed and balances resource requirements across hardware.Specifically, micro-CT scans of resin cast rat cerebral vasculature are segmented to remove reconstruction artifacts and are imported to instantiate nascent endothelial cells in a homogeneous three-dimensional grid that represents the area over which the simulation is performed.A growth factor source is added, and simulation of steady production and diffusion of the vascular endothelial growth factor (VEGF) is performed on the GPU using the NVidia Compute Unified Device Architecture (CUDA) programming platform.Motion of individual endothelial cells is then tracked over the lifetime of the simulation towards the source of growth factor, incorporating both sprouting and anastamosis events.
Angiogenesis, the formation of de novo blood vessels, has been implicated in many human diseases, including cancer, diabetes, and in recent years, neurodegenerative diseases. Study of the formation of microvasculature, therefore, has important implications in diagnosis and treatment of diseases. This paper describes how we combine a CPU-based Cellular Potts model of sprouting angiogenesis in three dimensions with medical imaging techniques and fluid dynamics equations to create an individual-based GPU accelerated angiogenesis simulation. The use of GPU, optimized for fast, highly parallel mathematical operations, provides an increase in simulation speed and balances resource requirements across hardware. Specifically, micro-CT scans of resin cast rat cerebral vasculature are segmented and imported to instantiate nascent endothelial cells in a homogeneous three-dimensional grid representing the area over which the simulation is performed. A growth factor source is added, and simulation of steady production and diffusion of the vascular endothelial growth factor (VEGF) is performed on the GPU on the NVidia Compute Unified Device Architecture (CUDA) programming platform. Motion of individual endothelial cells is then tracked over the lifetime of the simulation towards the source of growth factor.
This work details the results of a face authentication test (FAT2004) (http://www.ee.surrey.ac.uk/banca/icpr2004) held in conjunction with the 17th International Conference on Pattern Recognition. The contest was held on the publicly available BANCA database (http://www.ee.surrey.ac.uk/banca) according to a defined protocol (E. Bailly-Bailliere et al., June 2003). The competition also had a sequestered part in which institutions had to submit their algorithms for independent testing. 13 different verification algorithms from 10 institutions submitted results. Also, a standard set of face recognition software packages from the Internet (http://www.cs.colostate.edu/evalfacerec) were used to provide a baseline performance measure.
We present a fusion of Gabor feature based support vector machine (SVM) classifiers for face verification. 40 wavelets are used in parallel to extract features for face representation. These 40 feature extracted vectors are first projected onto the corresponding Principal Component Analysis (PCA) subspaces, and then fed into 40 SVMs for classification and fusion. No downsample is used. A publicly available FRAV2D face database with 4 different kinds of tests, each with 4 images per person, has been used to test our algorithm, considering frontal views, images with gestures, occlusions and changes of illumination. Compared to three baseline methods developed in literature, i.e. PCA, feature-based Gabor PCA and downsampled Gabor PCA, the proposed algorithm achieved the best results in the neutral expression and occlusion experiments. Compared to a downsampled Gabor PCA method, our algorithm also obtained similar error rates with a lower feature dimension.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)