Visual place recognition for persistent robot navigation in changing environments

This thesis demonstrates that robots can learn about how the world changes, and can use this information to recognise where they are, even when the appearance of the environment has changed a great deal. The ability to localise in highly dynamic environments using vision only is a key tool for achieving long-term, autonomous navigation in unstructured outdoor environments. The proposed learning algorithms are designed to be unsupervised, and can be generated by the robot online in response to its observations of the world, without requiring information from a human operator or other external source.

Visual rehabilitation in low-moderate keratoconus : intracorneal ring segment implantation followed by same-day topography-guided photorefractive keratectomy and collagen cross linking

AIM: To present the results of same-day topography-guided photorefractive keratectomy (TG-PRK) and corneal collagen crosslinking (CXL) after previous intrastromal corneal ring segment (ISCR) implantation for keratoconus. METHODS: An experimental clinical study on twenty-one eyes of 19 patients aged, 27.1±6.6 years (range: 19 – 43 years), with low to moderate keratoconus who were selected to undergo customized TG-PRK immediately followed by same-day CXL, 9 months after ISCR implantation in a university ophthalmology clinic. Refraction, uncorrected (UDVA) and corrected distance visual acuities (CDVA), keratometry (K) values, central corneal thickness (CCT) and coma were assessed 3 months after TG/PRK and CXL. RESULTS: After TG-PRK/CXL: the mean UDVA (logMAR) improved significantly from 0.66±0.41 to 0.20±0.25 (P<0.05); K flat value decreased from: 48.44±3.66 D to 43.71±1.95 D; K steep value decreased from 45.61±2.40 D to 41.56±2.05D; K average also decreased from 42.42±2.07 D to 47.00±2.66 D (P<0.05 for all). The mean sphere and cylinder decreased significantly post-surgery from, -3.10±2.99 D to -0.11±0.93 D and from, -3.68±1.53 to -1.11±0.75D respectively, while the CDVA, CCT and coma showed no significant changes. Compared to post-ISCR, significant reductions (P ˂ 0.05 or all) in all K-values, sphere and cylinder were observed after TG-PRK/CXL. CONCLUSION: Same-day combined topography-guided PRK and corneal crosslinking following placement of ICRS is a safe and potentially effective option in treating low-moderate keratoconus. It significantly improved all visual acuity, reduced keratometry, sphere and astigmatism, but caused no change in central corneal thickness and coma.

Visual field coordinates of pupillary circular axis and optical axis

Purpose: We term the visual field position from which the pupil appears most nearly circular as the pupillary circular axis (PCAx). The aim was to determine and compare the horizontal and vertical co-ordinates of the PCAx and optical axis from pupil shape and refraction information for only the horizontal meridian of the visual field. Method: The PCAx was determined from the changes with visual field angle in the ellipticity and orientation of pupil images out to ±90° from fixation along the horizontal meridian for the right eyes of 30 people. This axis was compared with the optical axis determined from the changes in the astigmatic components of the refractions for field angles out to ±35° in the same meridian. Results: The mean estimated horizontal and vertical field coordinates of the PCAx were (‒5.3±1.9°, ‒3.2±1.5°) compared with (‒4.8±5.1°, ‒1.5±3.4°) for the optical axis. The vertical co-ordinates of the two axes were just significantly different (p =0.03) but there was no significant correlation between them. Only the horizontal coordinate of the PCAx was significantly related to the refraction in the group. Conclusion: On average, the PCAx is displaced from the line-of-sight by about the same angle as the optical axis but there is more inter-subject variation in the position of the optical axis. When modelling the optical performance of the eye, it appears reasonable to assume that the pupil is circular when viewed along the line-of-sight.

Visual narratives and counter-narratives in aged care : Centring the participant as photographer in social science research

As detailed by a number of scholars (Emmison & Smith, 2000, 2012; Harrison, 1996, 2002, 2004), photographs and the process of photographing can provide fertile ground for sociological investigation. Examining the production of photography can tell us much about inclusion/omission and power/knowledge in a variety of social settings. Recently, some researchers have begun to utilise the participatory action research methodology, PhotoVoice, where people take and share photographs as a means of communicating and advocating on a specific topic. While medical sociologists have used PhotoVoice to communicate the impacts of disease in vulnerable populations (eg Burles, 2010), little social research has been done that combines PhotoVoice and older persons. This is interesting given the world’s population is ageing and the general lack of research that examines what daily life is like for older people living in aged care (Timonen & O’Dwyer, 2009). In response, a recent project tracked 10 participants who recently transitioned into living in residential aged care (RAC). The project combined the use of PhotoVoice methodology with repeated in-depth interviews. Residents were asked to orally and visually describe the positives and negative aspects of their daily lives. In the first instance, they shared the use of a RAC owned camera and later had the opportunity to access a camera for their sole use. Photographic analysis emphasised the value of centring the participant as an autonomous photographer in social research. In the photographs captured on a shared use camera, the photographs tended to depict predominately positive life stories (e.g. weekly morning tea outings, social activities). In comparison, the photographs captured on the sole use camera also described intimate but everyday activities, spaces, objects and people that frequented in their daily lives. Shifting the responsibility of the camera and photography solely to the participants resulted in the residents disrupting conventions of ‘suitable’ subject matter to photograph (Harrison, 2004) and in doing so, provided a much richer insight into what daily life is like in aged care.

Long-range stereo visual odometry for unmanned aerial vehicles

This thesis explored the utility of long-range stereo visual odometry for application on Unmanned Aerial Vehicles. Novel parameterisations and initialisation routines were developed for the long-range case of stereo visual odometry and new optimisation techniques were implemented to improve the robustness of visual odometry in this difficult scenario. In doing so, the applications of stereo visual odometry were expanded and shown to perform adequately in situations that were previously unworkable.

Visual control for automated aircraft collision avoidance systems

This thesis presents a new vision-based decision and control strategy for automated aircraft collision avoidance that can be realistically applied to the See and Avoid problem. The effectiveness of the control strategy positions the research as a major contribution toward realising the simultaneous operation of manned and unmanned aircraft within civilian airspace. Key developments include novel classical and visual predictive control frameworks, and a performance evaluation technique aligned with existing aviation practise and applicable to autonomous systems. The overall approach is demonstrated through experimental results on a small multirotor unmanned aircraft, and through high fidelity probabilistic simulation studies.

A neuronal topography of visual and acoustic fear conditioning within the amygdala

The lateral amygdala (LA) receives information from auditory and visual sensory modalities, and uses this information to encode lasting memories that predict threat. One unresolved question about the amygdala is how multiple memories, derived from different sensory modalities, are organized at the level of neuronal ensembles. We previously showed that fear conditioning using an auditory conditioned stimulus (CS) was spatially allocated to a stable topography of neurons within the dorsolateral amygdala (LAd) (Bergstrom et al, 2011). Here, we asked how fear conditioning using a visual CS is topographically organized within the amygdala. To induce a lasting fear memory trace we paired either an auditory (2 khz, 55 dB, 20 s) or visual (1 Hz, 0.5 s on/0.5 s off, 35 lux, 20 s) CS with a mild foot shock unconditioned stimulus (0.6 mA, 0.5 s). To detect learning-induced plasticity in amygdala neurons, we used immunohistochemistry with an antibody for phosphorylated mitogen-activated protein kinase (pMAPK). Using a principal components analysis-based approach to extract and visualize spatial patterns, we uncovered two unique spatial patterns of activated neurons in the LA that were associated with auditory and visual fear conditioning. The first spatial pattern was specific to auditory cued fear conditioning and consisted of activated neurons topographically organized throughout the LAd and ventrolateral nuclei (LAvl) of the LA. The second spatial pattern overlapped for auditory and visual fear conditioning and was comprised of activated neurons located mainly within the LAvl. Overall, the density of pMAPK labeled cells throughout the LA was greatest in the auditory CS group, even though freezing in response to the visual and auditory CS was equivalent. There were no differences detected in the number of pMAPK activated neurons within the basal amygdala nuclei. Together, these results provide the first basic knowledge about the organizational structure of two different fear engrams within the amygdala and suggest they are dissociable at the level of neuronal ensembles within the LA

IMU aided 3D visual odometry for car-like vehicles

We present a method for calculating odome- try in three-dimensions for car-like ground ve- hicles with an Ackerman-like steering model. In our approach we use the information from a single camera to derive the odometry in the plane and fuse it with roll and pitch informa- tion derived from an on-board IMU to extend to three-dimensions, thus providing odometric altitude as well as traditional x and y transla- tion. We have mounted the odometry module on a standard Toyota Prado SUV and present results from a car-park environment as well as from an off-road track.

Parallel attentive visual tracking

The research reported here addresses the problem of detecting and tracking independently moving objects from a moving observer in real-time, using corners as object tokens. Corners are detected using the Harris corner detector, and local image-plane constraints are employed to solve the correspondence problem. The approach relaxes the restrictive static-world assumption conventionally made, and is therefore capable of tracking independently moving and deformable objects. Tracking is performed without the use of any 3-dimensional motion model. The technique is novel in that, unlike traditional feature-tracking algorithms where feature detection and tracking is carried out over the entire image-plane, here it is restricted to those areas most likely to contain-meaningful image structure. Two distinct types of instantiation regions are identified, these being the “focus-of-expansion” region and “border” regions of the image-plane. The size and location of these regions are defined from a combination of odometry information and a limited knowledge of the operating scenario. The algorithms developed have been tested on real image sequences taken from typical driving scenarios. Implementation of the algorithm using T800 Transputers has shown that near-linear speedups are achievable, and that real-time operation is possible (half-video rate has been achieved using 30 processing elements).

Attentive visual tracking

The research reported here addresses the problem of detecting and tracking independently moving objects from a moving observer in real time, using corners as object tokens. Local image-plane constraints are employed to solve the correspondence problem removing the need for a 3D motion model. The approach relaxes the restrictive static-world assumption conventionally made, and is therefore capable of tracking independently moving and deformable objects. The technique is novel in that feature detection and tracking is restricted to areas likely to contain meaningful image structure. Feature instantiation regions are defined from a combination of odometry informatin and a limited knowledge of the operating scenario. The algorithms developed have been tested on real image sequences taken from typical driving scenarios. Preliminary experiments on a parallel (transputer) architecture indication that real-time operation is achievable.

Enhancing the teaching and learning of mathematical visual images

The literacy demands of mathematics are very different to those in other subjects (Gough, 2007; O'Halloran, 2005; Quinnell, 2011; Rubenstein, 2007) and much has been written on the challenges that literacy in mathematics poses to learners (Abedi and Lord, 2001; Lowrie and Diezmann, 2007, 2009; Rubenstein, 2007). In particular, a diverse selection of visuals typifies the field of mathematics (Carter, Hipwell and Quinnell, 2012), placing unique literacy demands on learners. Such visuals include varied tables, graphs, diagrams and other representations, all of which are used to communicate information.

From ImageNet to mining: Adapting visual object detection with minimal supervision

This paper presents visual detection and classification of light vehicles and personnel on a mine site.We capitalise on the rapid advances of ConvNet based object recognition but highlight that a naive black box approach results in a significant number of false positives. In particular, the lack of domain specific training data and the unique landscape in a mine site causes a high rate of errors. We exploit the abundance of background-only images to train a k-means classifier to complement the ConvNet. Furthermore, localisation of objects of interest and a reduction in computation is enabled through region proposals. Our system is tested on over 10km of real mine site data and we were able to detect both light vehicles and personnel. We show that the introduction of our background model can reduce the false positive rate by an order of magnitude.