A hidden Markov model and relative entropy rate approach to vision-based dim target detection for UAV sense-and-avoid

Uninhabited aerial vehicles (UAVs) are a cutting-edge technology that is at the forefront of aviation/aerospace research and development worldwide. Many consider their current military and defence applications as just a token of their enormous potential. Unlocking and fully exploiting this potential will see UAVs in a multitude of civilian applications and routinely operating alongside piloted aircraft. The key to realising the full potential of UAVs lies in addressing a host of regulatory, public relation, and technological challenges never encountered be- fore. Aircraft collision avoidance is considered to be one of the most important issues to be addressed, given its safety critical nature. The collision avoidance problem can be roughly organised into three areas: 1) Sense; 2) Detect; and 3) Avoid. Sensing is concerned with obtaining accurate and reliable information about other aircraft in the air; detection involves identifying potential collision threats based on available information; avoidance deals with the formulation and execution of appropriate manoeuvres to maintain safe separation. This thesis tackles the detection aspect of collision avoidance, via the development of a target detection algorithm that is capable of real-time operation onboard a UAV platform. One of the key challenges of the detection problem is the need to provide early warning. This translates to detecting potential threats whilst they are still far away, when their presence is likely to be obscured and hidden by noise. Another important consideration is the choice of sensors to capture target information, which has implications for the design and practical implementation of the detection algorithm. The main contributions of the thesis are: 1) the proposal of a dim target detection algorithm combining image morphology and hidden Markov model (HMM) filtering approaches; 2) the novel use of relative entropy rate (RER) concepts for HMM filter design; 3) the characterisation of algorithm detection performance based on simulated data as well as real in-flight target image data; and 4) the demonstration of the proposed algorithm's capacity for real-time target detection. We also consider the extension of HMM filtering techniques and the application of RER concepts for target heading angle estimation. In this thesis we propose a computer-vision based detection solution, due to the commercial-off-the-shelf (COTS) availability of camera hardware and the hardware's relatively low cost, power, and size requirements. The proposed target detection algorithm adopts a two-stage processing paradigm that begins with an image enhancement pre-processing stage followed by a track-before-detect (TBD) temporal processing stage that has been shown to be effective in dim target detection. We compare the performance of two candidate morphological filters for the image pre-processing stage, and propose a multiple hidden Markov model (MHMM) filter for the TBD temporal processing stage. The role of the morphological pre-processing stage is to exploit the spatial features of potential collision threats, while the MHMM filter serves to exploit the temporal characteristics or dynamics. The problem of optimising our proposed MHMM filter has been examined in detail. Our investigation has produced a novel design process for the MHMM filter that exploits information theory and entropy related concepts. The filter design process is posed as a mini-max optimisation problem based on a joint RER cost criterion. We provide proof that this joint RER cost criterion provides a bound on the conditional mean estimate (CME) performance of our MHMM filter, and this in turn establishes a strong theoretical basis connecting our filter design process to filter performance. Through this connection we can intelligently compare and optimise candidate filter models at the design stage, rather than having to resort to time consuming Monte Carlo simulations to gauge the relative performance of candidate designs. Moreover, the underlying entropy concepts are not constrained to any particular model type. This suggests that the RER concepts established here may be generalised to provide a useful design criterion for multiple model filtering approaches outside the class of HMM filters. In this thesis we also evaluate the performance of our proposed target detection algorithm under realistic operation conditions, and give consideration to the practical deployment of the detection algorithm onboard a UAV platform. Two fixed-wing UAVs were engaged to recreate various collision-course scenarios to capture highly realistic vision (from an onboard camera perspective) of the moments leading up to a collision. Based on this collected data, our proposed detection approach was able to detect targets out to distances ranging from about 400m to 900m. These distances, (with some assumptions about closing speeds and aircraft trajectories) translate to an advanced warning ahead of impact that approaches the 12.5 second response time recommended for human pilots. Furthermore, readily available graphic processing unit (GPU) based hardware is exploited for its parallel computing capabilities to demonstrate the practical feasibility of the proposed target detection algorithm. A prototype hardware-in- the-loop system has been found to be capable of achieving data processing rates sufficient for real-time operation. There is also scope for further improvement in performance through code optimisations. Overall, our proposed image-based target detection algorithm offers UAVs a cost-effective real-time target detection capability that is a step forward in ad- dressing the collision avoidance issue that is currently one of the most significant obstacles preventing widespread civilian applications of uninhabited aircraft. We also highlight that the algorithm development process has led to the discovery of a powerful multiple HMM filtering approach and a novel RER-based multiple filter design process. The utility of our multiple HMM filtering approach and RER concepts, however, extend beyond the target detection problem. This is demonstrated by our application of HMM filters and RER concepts to a heading angle estimation problem.

Two main challenges in service description : Web service tunnel vision and semantic myopia

Through media such as newspapers, letterbox flyers, corporate brochures and television we are regularly confronted with descriptions for conventional (bricks 'n' mortar style) services. These representations vary in the terminology utilised, the depth of the description, the aspects of the service that are characterised and their applicability to candidate service requestors. Existing service catalogues (such as the Yellow Pages) provide little relief for service requestors from the burdensome task of discovering, comparing and substituting services. Add to this environment the rapidly evolving area of web services with its associated surfeit of standards, and the result is a considerably fragmented approach to the description of services. It leaves the reality of the Semantic Web somewhat clouded. --------- Let's consider service description briefly, before discussing our concerns with existing approaches to description. The act of describing is performed prior to advertising. This simple fact provides an interesting paradox as services cannot be described exactly before advertisement. This doesn't mean they can't be described comprehensively. By "exactly", we are referring to the fact that context provided by a service requestor (and their service needs) will alter the description of the service that is presented to the discoverer. For example, a service provider who operates a cinema wants to describe the price of their service. Let's say the advertised price is $15. They also want to state that a pensioner discount and a student discount is available which provides a 50% discount. A customer (i.e. service requestor) uses the cinema web site to purchase tickets online. They find the movie of their choice at a time that suits. However, its not until some context is provided by the requestor that the exact price is determined. The requestor might state that they are a pensioner. The same is applicable for a service requestor who purchases multiple tickets perhaps on behalf of other people. The disconnect between when the service is described and when a requestor provides context introduces challenges to the description process. A service provider would be ill-advised to offer independent descriptions that represent all the permutations possible for a single service. The descriptive effort would be prohibitive.

Developing a vision for an nD modelling tool

The design of a building is a complicated process, having to formulate diverse components through unique tasks involving different personalities and organisations in order to satisfy multi-faceted client requirements. To do this successfully, the project team must encapsulate an integrated design that accommodates various social, economic and legislative factors. Therefore, in this era of increasing global competition integrated design has been increasingly recognised as a solution to deliver value to clients.----- The ‘From 3D to nD modelling’ project at the University of Salford aims to support integrated design; to enable and equip the design and construction industry with a tool that allows users to create, share, contemplate and apply knowledge from multiple perspectives of user requirements (accessibility, maintainability, sustainability, acoustics, crime, energy simulation, scheduling, costing etc.). Thus taking the concept of 3-dimensional computer modelling of the built environment to an almost infinite number of dimensions, to cope with whole-life construction and asset management issues in the design of modern buildings. This paper reports on the development of a vision for how integrated environments that will allow nD-enabled construction and asset management to be undertaken. The project is funded by a four-year platform grant from the Engineering and Physical Sciences Research Council (EPSRC) in the UK; thus awarded to a multi-disciplinary research team, to enable flexibility in the research strategy and to produce leading innovation. This paper reports on the development of a business process and IT vision for how integrated environments will allow nD-enabled construction and asset management to be undertaken. It further develops many of the key issues of a future vision arising from previous CIB W78 conferences.

Mapping of indoor environments by robots using low-cost vision sensors

For robots to operate in human environments they must be able to make their own maps because it is unrealistic to expect a user to enter a map into the robot’s memory; existing floorplans are often incorrect; and human environments tend to change. Traditionally robots have used sonar, infra-red or laser range finders to perform the mapping task. Digital cameras have become very cheap in recent years and they have opened up new possibilities as a sensor for robot perception. Any robot that must interact with humans can reasonably be expected to have a camera for tasks such as face recognition, so it makes sense to also use the camera for navigation. Cameras have advantages over other sensors such as colour information (not available with any other sensor), better immunity to noise (compared to sonar), and not being restricted to operating in a plane (like laser range finders). However, there are disadvantages too, with the principal one being the effect of perspective. This research investigated ways to use a single colour camera as a range sensor to guide an autonomous robot and allow it to build a map of its environment, a process referred to as Simultaneous Localization and Mapping (SLAM). An experimental system was built using a robot controlled via a wireless network connection. Using the on-board camera as the only sensor, the robot successfully explored and mapped indoor office environments. The quality of the resulting maps is comparable to those that have been reported in the literature for sonar or infra-red sensors. Although the maps are not as accurate as ones created with a laser range finder, the solution using a camera is significantly cheaper and is more appropriate for toys and early domestic robots.

Topological SLAM Using Fast Vision Techniques

In this paper we propose a method for vision only topological simultaneous localisation and mapping (SLAM). Our approach does not use motion or odometric information but a sequence of colour histograms from visited places. In particular, we address the perceptual aliasing problem which occurs using external observations only in topological navigation. We propose a Bayesian inference method to incrementally build a topological map by inferring spatial relations from the sequence of observations while simultaneously estimating the robot's location. The algorithm aims to build a small map which is consistent with local adjacency information extracted from the sequence measurements. Local adjacency information is incorporated to disambiguate places which otherwise would appear to be the same. Experiments in an indoor environment show that the proposed technique is capable of dealing with perceptual aliasing using visual observations only and successfully performs topological SLAM.