About me
I am a lecturer in the Department of Mechanical and Electrical Engineering at Massey University in Palmerston North, New Zealand. My research interests are mainly in colour and hyper-spectral image processing, unsupervised machine learning and human vision.<.

CV (English) last update: Nov. 2018
CV (French) last update: Nov. 2018

s.lemoan @ massey.ac.nz
Massey University profile
ORCID: 0000-0003-4713-2732

Linkedin    Researchgate    Scholar

Teaching & Supervision

I am currently involved in the following courses:

[281.756] Image and Video Processing [notes on compression]
[281.755] Digital Signal Processing
[159.270] Hardware-Oriented Programming
[281.353] Control Engineering [notes on fuzzy logic]
[228.311] Design with Constraints (3rd year project)
[228.271] Mathematics 2
[281.281] Analogue Electronics Systems



I am also supervising these student projects:

[PhD] Yuan Chang: Low-cost hyperspectral imaging (co-supervised by Prof. Donald Bailey)
[PhD] Sowmya Kasturi: Imaging and modelling of time/temperature integrators in foods undergoing microwave assisted thermal processing (co-supervised by Prof. Donald Bailey, project sponsored by FIET)
[PhD] Matthew Lofroth: Development of a Micro Manipulation Tool for Single Cell Analysis (with Dr Ebubekir Avci, project sponsored by Massey University)
[PhD] Saran Chowdary: Switched Reluctance Machine for Electric Vehicles application (with A/Prof. Ibrahim Al-Bahadly, project sponsored by the Callaghan Institute, in collaboration with Zero Emission Vehicles)
[ME] Hamish Bradley: Image registration for low-dose X-ray imaging (project sponsored by Volkswagen, in collaboration with the University of Stuttgart)
[4th yr project] Liam Millar: Data augmentation for deep learning-based leaf plant segmentation (project sponsored by Biolumic)
[4th yr project] Ben Matthews: Hyperspectral sensing-based classification of UV-treated seeds (project sponsored by Biolumic)
[4th yr project] Emily Aull: Colour sensing-based estimation of the pigment profile of spirulina (project sponsored by Tahi Spirulina)
[4th yr project] Ipeleng Motsatsi: Deep learning-based single image spectral super-resolution

Previously (see CV for full list):

[ME] Yuan Chang: Lens distortion correction with the Hough Transform (with Prof. Donald Bailey)
received a best student paper award at IVCNZ 2017
[4th yr project] Daniel de Waal: Low-cost control of ambient light chromaticity (with Dr. Huub Baker, sponsored by Fifth Season Design)
[4th yr project] Tom Biggs: Non-invasive milk level measurement in silos (sponsored by Levno)
[4th yr project] Liam Potts: Waterflow monitoring for farm management (sponsored by Levno)
[4th yr project] Jim Harvey: Exploting gradual change blindness for video coding
[4th yr project] Dylan Reid: Machine learning for prediction of eye movements
[4th yr project] Sheng Wang: RGB imaging-based meat quality measurement
[internship] Pierre Halle: Feature extraction for hyperspectral images (with Dr. Claude Cariou @ Universtiy of Rennes 1, France) invited for technical keynote presentation at IVCNZ 2017

Research


Spectral image analysis
Image quality assessment
Change blindness

Spectral image analysis


We can infer a great deal of information about an object by measuring how it interacts with light. For example, the colour of a banana is often a good predictor of its ripeness, and we can use this fact to decide when to harvest or to purchase them. However, colour is subjective (remember the black and blue dress?). It is not a physical measurement like distance or weight, it depends on psychophysical factors such as memory. Furthermore, colour is but a coarse representation of the light reflected by an object, just like a low-resolution picture would be a coarse representation of the spatial arrangement of objects within a scene. Fine details of the ligh/object interaction are completely discarded.
Just as increasing the number of pixels to have a "sharper" image, one can also increase the number of spectral bands from greyscale (1 band) or colour (3 bands) to multi-spectral (from 4 bands, typically only in the visible range 400-700 nm) and hyper-spectral (typically several hundreds of bands, mostly in the infrared range 700+ nm). Be it multi or hyper, spectral imaging does not just mimick the subjective sensation of colour, it measures the physical properties of pixels and objects in the scene. The figure to the left depicts how a vast majority of digital cameras work: three sensors capture electromagnetic energy in different parts of the visible spectrum of light, roughly corresponding to red, green and blue wavelength ranges. Eventually, the combination of all three values allows to reproduce color, for example on a display.


This figure illustrates how spectral capture works. Instead of capturing electromagnetic energy over three large ranges of visible wavelengths, spectral technology work at a higher spectral resolution and can measure energy in up to thousands of small ranges, including of non-visible lights such as ultra-violets (below 380nm) or infra-red (over 740nm).


Spectral imaging is used in a variety of applications such as cultural heritage (e.g. for digital archiving or art reproduction), remote sensing (e.g. for agricultural or environmental studies), but also medicine (e.g. skin analysis) or biometrics (e.g. vein patterns recognition). Spectral imaging produces datasets which are substantially larger than colour imaging, which results in certain challenges regarding computational effort, but also makes it difficult to extract information from it for classification (cf the "curse of dimensionality") or visualisation purposes.

Image quality assessment


In this figure, an observer is asked to rate the quality of a reproduced image, with respect to an undistorted reference.
The purpose of objective Image Quality Assessment is to mimic this subjective ratings based on the extraction and comparison of meaningful image features.




Change blindness


We see far less than we think we do: despite our impression of a richly detailed visual world, our perception is surprisingly limited. Consider this example of the "Spot the difference" game:


We found that most people initially believe that the two pictures are absolutely identical. This is due to change blindness (CB), a shortcoming of our visual system that is caused by limits of attention and memory. The proposed project aims to create a predictive model of change blindness in natural scenes. It will allow us to harness CB when it can help reduce file size and computational bandwidth (in image/video compression and computer graphics) and to avoid it when it can lead to errors and accidents (in car traffic, process plants, planes and even submarines). With the advent of streaming media, the increasing popularity of virtual and augmented reality, as well as the coming of low-cost eye-tracking and brain-computer interfacing, a deep understanding of visual perception has become both possible and crucial.
The study of CB and the factors responsible for it is a very important step towards that objective. We know that CB can be induced in different ways, and that it depends mostly on stimulus complexity, age and experience. It cannot be predicted by eye movements alone, and the current consensus is that CB emanates from a problem of encoding and comparison in visual working memory, specifically in the parietal cortex. However, there have been only a few attempts at creating a computational model able to automatically predict change detection performance in natural scenes. Existing approaches are ad hoc and based on bottom-up salience, which is known to be insufficient. A more elaborated model that accounts for individual differences to predict visual scanpaths and internal representations is needed, but the idiosyncrasy of CB and the paucity of available reference data have made it difficult to design and validate such model, particularly for applications related to visual quality.

Publications



Journal papers


Steven Le Moan and Claude Cariou "Minimax Bridgeness-based Clustering for Hyperspectral Data"
(submitted to) Remote Sensing, MDPI, 2020.

Donald Bailey, Yuan Chang and Steven Le Moan, "Analysing Arbitrary Curves from the Line Hough Transform"
(submitted to) Imaging, MDPI, 2020.

Liam Potts, Steven Le Moan and Gourab Sen Gupta, "Water Flow Monitoring for Fault Detection on Large Dairy Farms"
(submitted to) Agricultural Water Management, Elsevier, 2019.

Steven Le Moan and Marius Pedersen, "A Three-Feature Model to Predict Colour Change Blindness"
Vision, MDPI, 2019. [Open Access]

Steven Le Moan, Ivar Farup and Jana Blahová, "Towards Exploiting Visual Change Blindness for Image Processing"
Journal of Visual Communication and Image Representation, Elsevier, 2018. [pdf]

Steven Le Moan, Tejas Madan Tanksale, Roman Byshko and Philipp Urban, "An Observer-Metamerism Sensitivity Index for Electronic Displays"
Journal of the Society for Information Display, SID, 2017. (Fraunhofer IGD Best Paper Award) [pdf]

Jessica El-Khoury, Steven Le Moan, Jean-Baptiste Thomas and Alamin Mansouri, "Color and Sharpness Assessment of Single Image Dehazing"
Multimedia Tools and Applications, Springer, 2017. [pdf]

Steven Le Moan, Jana Blahová, Philipp Urban and Ole Norberg, "Five Dimensions for Spectral Colour Management"
Journal of Imaging Science and Technology, IS&T, vol. 60, no. 6, pp. 60501-1-60501-9(9), 2016. [pdf]

Steven Le Moan and Philipp Urban, "Image-Difference Prediction: From Color to Spectral"
IEEE Transactions on Image Processing, vol. 23, no. 5, pp. 2058-2068, 2014. [pdf] [supplementary material]

Steven Le Moan, Alamin Mansouri, Jon Yngve Hardeberg and Yvon Voisin, "Saliency for Spectral Image Analysis"
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 6, no. 6, pp. 2472-2479, 2013. [pdf][supplementary material]

Steven Le Moan, Alamin Mansouri, Yvon Voisin and Jon Yngve Hardeberg, "A Constrained Band Selection Method Based on Information Measures for Spectral Image Color Visualization"
IEEE Transactions on Geoscience and Remote Sensing, vol. 49, no. 12(2), pp. 5104-5115, 2011. [pdf]

Claude Cariou, Kacem Chehdi and Steven Le Moan,"Bandclust: An Unsupervised Band Reduction Method for Hyperspectral Remote Sensing"
IEEE Geoscience and Remote Sensing Letters, vol. 8, no. 3, pp. 565-569, 2010. [pdf]


Conference papers

(listed by year)

2020


Steven Le Moan, Virgil Tourneur and Gabor Kereszturi "Visualisation of Hyperspectral Remote Sensing Data via Colour Compositing: a Subjective Study"
(submitted to) International Conference on Image Processing, 2020. IEEE

Steven Le Moan, Marius Pedersen and Aladine Chetouani "High-Level Visual Masking of Image Compression Aretefacts"
(submitted to) International Conference on Image Processing, 2020. IEEE

Jim Harvey and Steven Le Moan "Gradual Chroma Reduction and High-level Visual Masking in Videos"
(submitted to) International Conference on Image Processing, 2020. IEEE

Aladine Chetouani, Marius Pedersen and Steven Le Moan "Prediction of Chromatic Visual Masking with Deep Learning"
(submitted to) International Conference on Image Processing, 2020. IEEE


2019


Claude Cariou, Kacem Chehdi and Steven Le Moan "Improved Nearest Neighbor Density-Based Clustering Techniques with Application to Hyperspectral Images"
ICASSP, 2019. IEEE

Hamish Bradley, Donald Bailey, Steven Le Moan, Peter Gaenz and Sven Simon "Sub-pixel Registration Techniques for X-ray Phase Contrast Imaging"
Image and Vision Computing New Zealand, 2019. IEEE proceedings

Yuan Chang, Steven Le Moan and Donald Bailey "RGB Imaging Based Estimation of Leaf Chlorophyll Content"
Image and Vision Computing New Zealand, 2019. IEEE proceedings

Yuan Chang, Donald Bailey and Steven Le Moan "The shape of patterns tells more: Using two-dimensional Hough transform to detect circles"
Asian Conference on Pattern Recognition, 2019. Springer

Claude Cariou, Kacem Chehdi and Steven Le Moan "GWENN-SS: a simple semi-supervised nearest-neighbor density-based classification method with application to hyperspectral images"
Remote Sensing, 2019. SPIE

Steven Le Moan and Marius Pedersen "Subjective Image Fidelity Assessment: Effect of the Spatial Distance Between Stimuli"
International Conference on Image Processing, 2019. IEEE

2018


Donald Bailey, Yuan Chang and Steven Le Moan "Lens Distortion Self-Calibration using the Hough Transform"
International Conference on Field-Programmable Technology, 2018. [pdf]

Steven Le Moan, Claude Cariou, "Parameter-Free Density Estimation for Hyperspectral Image Clustering"
Image and Vision Computing New Zealand, 2018, IEEE proceedings.(Voted second best paper) [pdf] [presentation]

Steven Le Moan, Marius Pedersen, "Measuring the Effect of High-level Visual Masking in Subjective Image Quality Assessment with Priming"
International Conference on Image Processing, 2018, IEEE. [pdf]

2017


Steven Le Moan, "Blind Classification for Remote Sensing Hyperspectral Images"
PROTINUS Workshop, 2017, University of Auckland.(Invited talk)

Steven Le Moan, "Can Image Quality Features Predict Visual Change Blindness?"
Image and Vision Computing New Zealand, 2017, IEEE proceedings. [pdf]

Pierre Hallé, Steven Le Moan, Claude Cariou, "Towards a Completely Blind Classifier for Hyperspectral Images"
Image and Vision Computing New Zealand, 2017, IEEE proceedings. (Keynote) [pdf]

Yuan Chang, Donald Bailey, Steven Le Moan, "Lens Distortion Correction by Analysing Peak Shape in Hough Transform Space"
Image and Vision Computing New Zealand, 2017, IEEE proceedings. (Best Student Paper Award) [pdf]

Steven Le Moan, Donald Bailey, "Comparison of Machine Learning-Based Feature Pooling Strategies for Colour Image Fidelity Assessment"
Image and Vision Computing New Zealand, 2017, IEEE proceedings. [pdf]

Daniel de Waal, Steven Le Moan, Huub Bakker, "Adaptive Control of the Colour Rendering of LED Lighting with an RGBW Sensor"
Image and Vision Computing New Zealand, 2017, IEEE proceedings. [pdf]

Steven Le Moan, Marius Pedersen, "Evidence of change blindness in subjective image fidelity assessment"
International Conference on Image Processing, 2017, IEEE [pdf]

2016


Steven Le Moan, "How to predict lightness variations from one illuminant to another?"
in Image and Vision Computing New Zealand, 2016, IEEE proceedings. [pdf]

Steven Le Moan, Marius Pedersen, Ivar Farup, Jana Blahová, "The Influence of Short-Term Memory in Subjective Image Quality Assessment"
in International Conference on Image Processing, 2016, IEEE. [pdf]

Steven Le Moan, "Quality Assessment of Spectral Reproductions: the Camera's Perspective"
in International Conference on Image Analysis and Recognition, 2016, Springer. [pdf]

2015


Steven Le Moan and Ivar Farup, "Exploiting Change Blindness for Image Compression"
in 11th International Conference on Signal, Image, Technology and Internet Based Systems, Bangkok, November 2015, IEEE proceedings. (Best Paper Award) [pdf]

Steven Le Moan and Ludovic Gustafsson Coppel, "Perceived Quality of Printed Images on Fluorescing Substrates under Various Illuminations"
in Proceedings of the 16th International Symposium on Multispectral Color Science, AIC 2015 Mid-term meeting, Tokyo, May 2015, Color Science Association of Japan. [pdf]

Steven Le Moan and Philipp Urban, "Evaluating the Multi-Scale iCID Metric"
in Image Quality and System Performance XII, Mohamed-Chaker Larabi and Sophie Triantaphillidou, Eds., San Francisco, CA, USA, February 2015, vol. 9396, p. 38, IS&T/SPIE. [pdf][supplementary material]

Steven Le Moan, Sony George, Marius Pedersen, Jana Blahová, and Jon Yngve Hardeberg, "A Database for Spectral Image Quality"
in Image Quality and System Performance XII, Mohamed-Chaker Larabi and Sophie Triantaphillidou, Eds., San Francisco, CA, USA, February 2015, vol. 9396, p. 25, IS&T/SPIE. [pdf]

2014


Steven Le Moan and Philipp Urban, "Spectral Printing with a CMYKRGB printer: a Closer Look"
in Proceedings of the 22nd Color and Imaging Conference, Boston, MA, USA, November 2014, pp. 131-135, IS&T. [pdf]

Ludovic Gustafsson Coppel, Steven Le Moan, Paula Zitinski Elias, Radovan Slavuj and Jon Yngve Hardeberg, "Next Generation Printing - Towards Spectral Proofing"
in Advances in Printing and Media Technology - Print and Media Research for the Benefit of Industry and Society, 41st International IARIGAI Conference, Swansea, UK, September 2014. [pdf]

Steven Le Moan and Philipp Urban, "A New Connection Space for Low-Dimensional Spectral Color Management"
in Measuring, Modeling, and Reproducing Material Appearance, Maria V. Ortiz Segovia, Philipp Urban, and Jan P. Allebach, Eds., San Francisco, CA, USA, February 2014, vol. 9018, p. 12, IS&T/SPIE. [pdf]


2013


Steven Le Moan and Philipp Urban, "Image quality and change of illuminant: An information-theoretic evaluation"
in Proceedings of the 21st Color and Imaging Conference, Albuquerque, NM, USA, November 2013, pp. 102-107, IS&T. [pdf]

Steven Le Moan and Philipp Urban, "Evaluating the perceived quality of spectral images"
in Image Processing, 20th International Conference on, Melbourne, Australia, September 2013, pp. 2024-2028, IEEE. [pdf]

Jana Blahová, Steven Le Moan, and Philipp Urban, "The impact of illumination on the perceived quality of spectral reproductions"
in Proceedings of the 19th Color Image Processing Workshop, Berlin, Germany, September 2013, pp. 93-101, German Color Group. [pdf]


2012


Heidi Sårheim, Steven Le Moan, Joni Nersveen, Jon Yngve Hardeberg, and Andreas Poppe, "Eye disease simulator - How do we see the world when vision is failing?"
in Universal Design - Public space: Inspire, Challenge, and Empower, Oslo, Norway, June 2012. [abstract]

Steven Le Moan, Ferdinand Deger, Alamin Mansouri, Yvon Voisin, and Jon Yngve Hardeberg, "Salient pixels and dimensionality reduction for display of multi/hyperspectral images"
in Image and Signal Processing, 5th International Conference on, Abderrahim Elmoataz, Driss Mammass, Olivier Lezoray, Fathallah Nouboud, and Driss Aboutajdine, Eds., Agadir, Morocco, June 2012, vol. 7340 of Lecture Notes in Computer Science, pp. 9-16, Springer. [pdf]


2011


Steven Le Moan, Alamin Mansouri, Jon Yngve Hardeberg, and Yvon Voisin, "Saliency-based band selection for spectral image visualization"
in Proceedings of the 19th Color Imaging Conference, San Jose, CA, USA, November 2011, pp. 363-368, IS&T. [pdf]

Steven Le Moan, Alamin Mansouri, Jon Yngve Hardeberg, and Yvon Voisin, "Visualization of spectral images: a comparative study"
in Proceedings of the 6th Gjøvik Color Imaging Symposium, September 2011, Gjøvik University College. [pdf]

Steven Le Moan, Alamin Mansouri, Yvon Voisin, and Jon Yngve Hardeberg, "Sélection de bandes pour la visualisation d'images spectrales: une approche basée sur l'étude de saillance"
in 23ème Colloque GRETSI, Bordeaux, France, September 2011. [pdf]

Steven Le Moan, Alamin Mansouri, Yvon Voisin, and Jon Yngve Hardeberg, "Spatially variant dimensionality reduction for the visualization of multi/hyperspectral images"
in Image Analysis and Recognition, International Conference on, Mohamed Kamel and Aurelio Campilho, Eds., Burnaby, BC, Canada, June 2011, vol. 6753 of Lecture Notes in Computer Science, pp. 375-384, Springer. [pdf]

Steven Le Moan, Alamin Mansouri, Yvon Voisin, and Jon Yngve Hardeberg, "Visualisation d'images spectrales : une méthode basée sur la perception humaine"
in 13ème ORASIS - Journées francophones des jeunes chercheurs en vision par ordinateur, Praz-sur-Arly, France, June 2011. [pdf]

Steven Le Moan, Alamin Mansouri, Jon Yngve Hardeberg, and Yvon Voisin, "Saliency in spectral images"
in Image Analysis, 17th Scandinavian Conference on, Anders Heyden and Fredrik Kahl, Eds., Ystad, Sweden, May 2011, vol. 6688 of Lecture Notes in Computer Science, pp. 114-123, Springer. [pdf]


2010


Steven Le Moan, Alamin Mansouri, Yvon Voisin, and Jon Yngve Hardeberg, "An efficient method for the visualization of spectral images based on a perception-oriented spectrum segmentation"
in Advances in Visual Computing, 6th International Symposium on, George Bebis, Richard Boyle, Bahram Parvin, Darko Koracin, Ronald Chung, Riad Hammoud, Muhammad Hussain, Tan Kar-Han, Roger Crawfis, Daniel Thalmann, David Kao, and Lisa Avila, Eds., Las Vegas, NV, USA, November 2010, vol. 6453 of Lecture Notes in Computer Science, pp. 361-370, Springer. [pdf]

Steven Le Moan, Alamin Mansouri, Jon Yngve Hardeberg, and Yvon Voisin, "A class-separability-based method for multi/hyperspectral image color visualization"
in Image Processing, 17th International Conference on, Hong Kong, September 2010, pp. 1321-1324, IEEE. [pdf]

Steven Le Moan, Alamin Mansouri, Yvon Voisin, and Jon Yngve Hardeberg, "Convex objects recognition and classification using spectral and morphological descriptors"
in Colour in Graphics, Imaging, and Vision, 5th European Conference on, Joensuu, Finland, June 2010, pp. 293-299, IS&T. [pdf]