No description available.

No description available.

The research described in this thesis involves the development of a video coding scheme which provides lossless Region Of Interest (ROl) coding specifically targeted toward the compression of thermal video sequences. In particular the scheme is designed to provide lossless coding of critical temperature data contained in thermal video sequences used within firefighting applications. High coding efficiency is achieved by the use of Discrete Cosine Transform (DCT) based lossy coding for non-critical areas of the video sequence. The research is divided into three key sections; an investigation into thermal video, the practical development of a novel scheme for selective lossless compression of thermal video containing critical temperature data, and the presentation and analysis of test results which demonstrate the effectiveness of the scheme. The scheme's multi-pass encoder stores data which enables the recovery of the losses which occur within critical thermal data as a result of the lossy DCT coding stage. This error recovery data is stored as codewords where each individual codeword is used to represent the error data for up to two pixels within the ROI. Efficient entropy coding of the error recovery data is achieved through a combination of a new adaptive reordering scheme followed by Run Length Coding (RLC) and Variable Length Coding (VLC). By storing these error recovery codes within the normally unused colour channels of a compressed video sequence, the compressed thermal video sequence can be transmitted as an MPEG-1, MPEG-2 and MPEG4 compliant bitstream. This novel storage mechanism allows the error recovery codes to be spatially mapped to the ROl data, and hence the scheme can benefit from the motion compensation facilities provided within the MPEG compliant bitstream. Decoding of the lossy video sequence can be carried out using any MPEG compliant decoder. Lossless ROl decoding can be carried out using the special decoder described in, and provided with, the thesis.

Poster exhibited at GPSC Student Showcase, February 24th, 2016, University of Arizona.

Different ways to measure human affective and deceptive reactions to stimulus have been developed. One method is a multimodal approach using web camera, thermal imaging camera and physiological sensors data to extract different features in the human face (verbal and non-verbal behavior) such as breathing rate, heart rate, face temperature, skin conductance, eye tracking, language analysis and facial expressions among others. Human subjects from different ages and ethnicity were exposed to two different experiments were they watched videos (affection recognition) and others answered an interview session (deception recognition). With the data collected from videos (thermal and visual), different regions of interest (ROI) of the face were selected as well as the whole picture. The ROI were determined based on the most sensitive parts of the face where larger changes of temperature or other physiological features are recorded. It was also analyzed the language (written and spoken) in order to obtain the verbal modalities. The data has been compared among the subjects to determine whether the deceptive and affective reactions of a person can be predicted using multimodal approach. From the multiple data obtained, a characterization of reactions is proposed when subjects are exposed to different stimulus, positive or negative, as well as deceptive behavior and later on recognize if the person is happy, sad, nervous, anxious, telling the truth, lying etc. Using the multimodal approach we were able to predict automatically, with higher accuracy than the baseline, affective and deceptive states of a person. In the affective state recognition, the classifier software differentiated affective state versus neutral state with 92.85% accuracy. Then it differentiated Positive State, Negative State and Neutral State with 57.14% accuracy. Additionally, it differentiated Positive State versus Negative State with 73.21% accuracy. Finally, the classifier was able to predict Deceptive State (people lying) and Non Deceptive State (people telling the truth) with 72.72% accuracy.

The objectives of this study were (1) to evaluate the use of digital infrared thermal imaging as a diagnostic tool for early detection of mastitis, (2) to test the usefulness of thermography in the quantative evaluation of mammary function in high and low milk producing cows and (3) to establish base-line measurements of thermographic patterns of the bovine claw. To investigate these objectives, three experiments were conducted to determine the link between thermal temperature gradients of the bovine mammary gland to level of milk production (high milk producers vs. low milk producers), early diagnosis of mastitis and lameness evaluation in dairy cattle. The first study investigated the relationship between level of milk production to udder conformation and thermal signatures of the udder gradients. The second study (case study) involved the use of DITI for diagnosis of mastitis in dairy cows under a heatstressed environment. Lastly, the third study was to establish base-line measurements in temperature gradients of regions of the bovine claw.

No description available.

This thesis reports a summary of the PhD programme for the assessment of person‘s emotional anxiety using Electro-optical technology. The thesis focuses mainly on the understanding of fundamental properties of physiological responses to emotional anxiety and how they can be captured by using Electro-optical (EO) imaging methods such as hyperspectral imaging (HSI) and thermal imaging (TI) techniques. The thesis summarises three main areas of work that have been undertaken by the author in the programme: (a) Experimental set up including HSI system and data acquisition software design and implementation, (b) fundamental understanding of physiological responses to emotional anxiety from the EO perspective and (c) the development of a novel remote sensing technique for the assessment of emotions without the requirement of base line information. One of our main results is to provide evidence to prove that the mean temperature in the periorbital region remains the same within 0.2°C during emotional anxiety. Furthermore, we have shown that it is the high temperature pixels within the periorbital, which increases in numbers by a huge amount after 2 minutes of the onset of anxiety. We have also developed techniques to allow the assessment anxiety without the need of base line information. The method has been tested using a sample size of about 40 subjects, and achieved promising result. Technologies for the remote sensing of heart beat rate has been in great demand, this study also involves the development of heart beat detection using TI system. Moreover, we have also attempted for the first time to sense glucose concentration from the blood sample in-vivo using HSI technique remotely.

The objective of this thesis was to evaluate the feasibility of three approaches to thermal imaging that used fiber-optics to decouple the imaging optic from a detector in order to form images without a direct line of sight. All images were formed using 2 m-long fibers, at a working distance of 19 cm, and the ability to detect temperatures between 100 °C to 200 °C. The three approaches were designated active, hybrid, and passive. The active approach involved scanning a single fiber in the image plane to formulate thermal images. Although images had sufficient field of view and spatial resolution, they were not formulated at a frequency of at least 1 Hz using either linear stages or a piezoelectric actuator for scanning. The hybrid approach involved scanning a bundle of fibers in the image plane to formulate a thermal image. Although it was possible to form individual images from fibers within a simple bundle and a seven-fiber bundle with sufficient field of view and spatial resolution, an image formulation frequency of at least 1 Hz was not achieved. While considered feasible, future work is required to identify a suitable motion platform, design a new bundle, and develop a method for stitching images from individual fibers into a single spatially-coherent image in order to achieve an image formulation frequency of at least 1 Hz. Lastly, the passive approach was evaluated in which images were formed using a fixed seven-fiber bundle for direct thermal imaging. The seven-fiber bundle produced images with sufficient field of view and an image formulation frequency on the order of kHz, but without sufficient spatial resolution. Additional work is required to determine the feasibility of designing a passive bundle for a specific spatial resolution, as well as, if overall costs are prohibitive. While no approach satisfied all functional requirements specified herein, evaluation of each approach and a potential solution are provided.

No description available.