Alan E. McKinnon

A model for measurement of noise in CCD digital-video cameras

This study presents a comprehensive measurement of CCD digital-video camera noise. Knowledge of noise detail within images or video streams allows for the development of more sophisticated algorithms for separating true image content from the noise generated in an image sensor. The robustness and performance of an image-processing algorithm is fundamentally limited by sensor noise. The individual noise sources present in CCD sensors are well understood, but there has been little literature on the development of a complete noise model for CCD digital-video cameras, incorporating the effects of quantization and demosaicing.

A Technique for Evaluation of CCD Video-Camera Noise

This paper presents a technique to identify and measure the prominent sources of sensor noise in commercially available charge-coupled device (CCD) video cameras by analysis of the output images. Noise fundamentally limits the distinguishable content in an image and can significantly reduce the robustness of an image processing application. Although sources of image sensor noise are well documented, there has been little work on the development of techniques to identify and quantify the types of noise present in CCD video-camera images. A comprehensive noise model for CCD cameras was used to evaluate the technique on a commercially available CCD video camera.

A model of magnesium metabolism in young sheep. Magnesium absorption and excretion

A model of Mg metabolism in sheep is proposed. It is based on standard Michaelis-Menten enzyme kinetics to describe the transport of Mg across the rumen wall and passive diffusion to describe the absorption of Mg in the hindgut. Factors known to have an effect on Mg metabolism in farm animals, namely the concentrations of K and Mg in the diet, and the physico-chemical conditions within the rumen as determined by the type of diet, are incorporated into the model. Consideration of the rumen as the only site of Mg absorption provided an inadequate mechanistic description of Mg metabolism in sheep. To ensure compatibility between predicted Mg absorption and recent independent data sets for Mg balances, it was necessary to include in the model aspects of Mg absorption that operate in the hindgut. The results from this model suggest that there is a need for a series of experiments to determine the important aspects of Mg transport in the hindgut of sheep. Mechanisms of homeostasis are discussed.

A model of magnesium metabolism in young sheep: transactions between plasma, cerebrospinal fluid and bone.

An extension of the previously proposed model of Mg metabolism (Robson et al. 1997) has been developed to consider the transactions of Mg that are associated with cerebrospinal fluid (CSF) and bone. The representation of the CSF as a single MG compartment with uptake from the plasma described by Michaelis-Menten kinetics gives very good agreement with published experiments. Analysis of the available information on resorption of Mg from adult bone indicated that this process makes a negligible contribution to Mg homeostasis and can be omitted from the model.

A desktop virtual reality application for chemical and process engineering education

A desktop Virtual Reality (VR) application of a skim milk powder process has been developed. The goal was to use this application as a learning resource to expose students to processing plant environments, which are becoming increasingly difficult to visit due to availability and safety reasons. The VR application comprises 360° panorama images of the milk powder process plant, process flow diagrams (PFD), piping and instrumentation diagrams (P&ID), 3D drawings and additional materials such as supplemental text, videos and animations. This paper describes the VR application as well as an evaluation of the application. The results of this study show that the application was easy to use and the users were satisfied with it. The design recommendations for the development of similar VR learning applications are also discussed in this paper.

Edge-based detection of sky regions in images for solar exposure prediction

A device for predicting the solar exposure at a location operates by gathering image data from that location with a known camera orientation. The image data is then processed to identify the sky regions and the solar exposure is predicted using a standard sun path model and tracing the rays from the sun through the processed images. Critical to the success of this technique is the image processing used to separate the sky from the rest of the image. This work is concerned with developing a technique which can do this for images taken under different weather conditions. The general approach to separate the sky from the rest of the image is to use the Canny edge detector and the morphology closing algorithm to find the regions in the image. The brightness and area of each region are then used to determine which regions are sky. The FloodFill algorithm is applied to identify all pixels in each sky region.

Noise-bound method for detecting shadow-free scene changes in image sequences

Many image processing applications are confounded by both sensor noise and cast shadows. All image sensors add noise to a captured image that can reduce algorithm sensitivity and performance, and global filters or fixed thresholds are often applied to limit their effects. Cast shadows can appear as scene changes and are difficult to adequately detect and remove from images and image sequences. We couple image-noise statistics with a dual-illumination shadow-detection algorithm to provide a novel color-based method for shadow-free scene-change detection whose performance is bound by metamerism and image noise, and has only one variable--the desired confidence interval for noise separation.

An investigation into noise-bound shadow detection and removal

Noise is an unavoidable contaminant in any non-trivial image. It is usually identified as a limiting factor in the performance of shadow-removal algorithms, but little is done to reduce its negative impact. The typical method to counter noise effects is to employ arbitrary or empirical thresholds somewhere inside the algorithm, with values chosen to maximize the shadow-removal performance. However these thresholds can be objectively calculated from the noise statistics for a particular pixel value. We present a method of shadow-removal whose internal parameters are adaptively set by noise statistics such that the algorithm is free of any empirically set threshold. Experiments indicate that the performance of the new method is approximately equivalent to that with an empirically-fixed threshold, though an area of improvement has been identified that could significantly boost the accuracy of the new method.

Pragmatic data modelling and design for end users

Many people are dependent on desktop end user tools such as spreadsheets and databases to manage their data. While they may have the technical skills to set up data repositories, many end users lack the analysis skills to design data models which reflect their often deceptively complex requirements. We advocate that a comprehensive data model should always be developed, with expert help, so that the end user can feel confident the subtleties of the data are fully understood. We then suggest that some pragmatic decisions can be made to simplify the model so that the end user can retain control over setting up and maintaining the application.

Expert system for flood management in Lake Manapouri

Flood water in Lake Manapouri is released according to strictly formulated flood rules. The Real-time Flood Assistant is an expert system which incorporates Lake Manapouri flood rules and the experience of the control room operators at Transpower NZ Ltd. to assist them in the release of flood water. The expert system is being developed in Level5 Object. The program is mainly an event-driven system based on forward chaining inferencing in conjunction with object-oriented methods. Backward chaining inferencing is used only at the beginning of the program to establish initial data. With the use of built-in clocks, the expert system runs in real-time, assisting the operators with the release decisions which are made about every 1 1/2 hours.<<ETX>>