published_or_final_version / Computer Science / Master / Master of Philosophy

published_or_final_version / Chinese / Master / Master of Philosophy

No description available.

In this thesis it is shown that several pattern recognition problems can be solved efficiently by exploiting the geometrical structure of the problems. The problems considered are in the area of clustering and classification. These problems are: (i) computing the diameter of a finite planar set, (ii) computing the maximum and minimum distance between two finite planar sets of points, (iii) testing for point inclusion in a convex polyhedron in d-dimensional space, and (iv) exact and inexact reference set thinning for the nearest neighbor decision rule. / Algorithms to solve the above problems are presented and analyzed for worst-case and average-case situations. These algorithms are implemented and experimentally compared with the existing algorithms. / In solving the above problems, a geometrical construct, known as the Voronoi diagram is used extensively. However, there exists no practical algorithm to construct the Voronoi diagram in d dimensional spaces when d > 2. In this thesis an efficient algorithm to construct the Voronoi diagram in d-space is presented.

No description available.

No description available.

No description available.

No description available.

This thesis addresses the problem of size invariant shape recognition based on scale transformation within modulated competition neural layer. In this thesis I will present the advantages of applying neural networks in pattern recognition and study how the traditional automatic target recognition fails to recognise known patterns due to size change, cluttered backgrounds and distortion. Within the thesis we will also discuss possible ways to overcome size variance and how the combining of Selective Attention Adaptive Resonance Theory makes the system capable of recognising images with size changes, distortion and in complex backgrounds. The model is constructed based on neurophysiology experiments in vision systems. The Neural Circuit Simulation studies undertaken demonstrate the effectiveness of the proposed model in recognising 2D objects in many non-ideal visual conditions. Despite size differences from the stored memory image, difficult visual environments, including severe distortion, the simulation results indicate the model can recognise the shape stored in memory from the simulated shapes. / From the research presented in this thesis, it is concluded that the use of attentional mechanisms can enhance artificial vision systems to cope with difficult visual conditions. It is shown that feed-forward-feedback interactions with synaptic modulation are a versatile and powerful mechanism for performing many useful functions such as gain control, filtering and selective processing in neural network based vision systems. / Thesis (MEng(ComputerSystemsEng))--University of South Australia, 2004.

The construction of surface models from sensor data is an important part of perceptive robotics. When the sensor data are obtained from fixed sensors the problem of occlusion arises. To overcome occlusion, sensors may be mounted on a robot that moves the sensors over the surface. In this thesis the sensors are single–point range finders. The range finders provide a set of sensor points, that is, the surface points detected by the sensors. The sets of sensor points obtained during the robot’s motion are used to construct a surface model. The surface model is used in turn in the computation of the robot’s motion, so surface modelling is performed on–line, that is, the surface model is constructed incrementally from the sensor points as they are obtained. A planar polyhedral surface model is used that is amenable to incremental surface modelling. The surface model consists of a set of model segments, where a neighbour relation allows model segments to share edges. Also sets of adjacent shared edges may form corner vertices. Techniques are presented for incrementally updating the surface model using sets of sensor points. Various model segment operations are employed to do this: model segments may be merged, fissures in model segment perimeters are filled, and shared edges and corner vertices may be formed. Details of these model segment operations are presented. The robot’s control point is moved over the surface model at a fixed distance. This keeps the sensors around the control point within sensing range of the surface, and keeps the control point from colliding with the surface. The remainder of the robot body is kept from colliding with the surface by using redundant degrees–of–freedom. The goal of surface modelling and surface following is to model as much of the surface as possible. The incomplete parts of the surface model (non–shared edges) indicate where sections of surface that have not been exposed to the robot’s sensors lie. The direction of the robot’s motion is chosen such that the robot’s control point is directed to non–shared edges, and then over the unexposed surface near the edge. These techniques have been implemented and results are presented for a variety of simulated robots combined with real range sensor data.