Guillaume Noyel

Morphological segmentation of hyperspectral images

The present paper develops a general methodology for the morphological segmentation of hyperspectral images, i.e., with an important number of channels. This approach, based on watershed, is composed of a spectral classification to obtain the markers and a vectorial gradient which gives the spatial information. Several alternative gradients are adapted to the different hyperspectral functions. Data reduction is performed either by Factor Analysis or by model fitting. Image segmentation is done on different spaces: factor space, parameters space, etc. On all these spaces the spatial/spectral segmentation approach is applied, leading to relevant results on the image.

Random germs and stochastic watershed for unsupervised multispectral image segmentation

This paper extends the use of stochastic watershed, recently introduced by Angulo and Jeulin [1], to unsupervised segmentation of multispectral images. Several probability density functions (pdf), derived from Monte Carlo simulations (M realizations of N random markers), are used as a gradient for segmentation: a weighted marginal pdf a vectorial pdf and a probabilistic gradient. These gradient-like functions are then segmented by a volume-based watershed algorithm to define the R largest regions. The various gradients are computed in multispectral image space and in factor image space, which gives the best segmentation. Results are presented on PLEIADES satellite simulated images.

Filtering, segmentation and region classification by hyperspectral mathematical morphology of DCE-MRI series for angiogenesis imaging

Segmenting dynamic contrast enhanced-MRI series of small animal, which are intrinsically noisy and low contrasted images with low resolution, is the aim of this paper. To do this, a segmentation method taking into account the temporal (spectral) and spatial information is presented on several series. The idea is to start from a temporal classification, and to build a probability density function of contours conditionally to this classification. Then, this function is segmented to find potentially tumorous areas. The method is presented on several series after a range normalization histogram in order to compare the series.

FAST COMPUTATION OF ALL PAIRS OF GEODESIC DISTANCES

Computing an array of all pairs of geodesic distances between the pixels of an image is time consuming. In the sequel, we introduce new methods exploiting the redundancy of geodesic propagations and compare them to an existing one. We show that our method in which the source point of geodesic propagations is chosen according to its minimum number of distances to the other points, improves the previous method up to 32 % and the naive method up to 50 % in terms of reduction of the number of operations.

Regionalized Random Germs by a Classification for Probabilistic Watershed Application: Angiogenesis Imaging Segmentation

New methods are presented to generate random germs regionalized by a previous classification in order to use probabilistic watershed on hyperspectral images. These germs are much more efficient than the standard uniform random germs.

A new spatio-spectral morphological segmentation for multi-spectral remote-sensing images

A general framework of spatio-spectral segmentation for multi-spectral images is introduced in this paper. The method is based on classification-driven stochastic watershed (WS) by Monte Carlo simulations, and it gives more regular and reliable contours than standard WS. The present approach is decomposed into several sequential steps. First, a dimensionality-reduction stage is performed using the factor-correspondence analysis method. In this context, a new way to select the factor axes (eigenvectors) according to their spatial information is introduced. Then, a spectral classification produces a spectral pre-segmentation of the image. Subsequently, a probability density function (pdf) of contours containing spatial and spectral information is estimated by simulation using a stochastic WS approach driven by the spectral classification. The pdf of the contours is finally segmented by a WS controlled by markers from a regularization of the initial classification.

MULTIVARIATE MATHEMATICAL MORPHOLOGY FOR DCE-MRI IMAGE ANALYSIS IN ANGIOGENESIS STUDIES

We propose a new computer aided detection framework for tumours acquired on DCE-MRI (Dynamic Contrast Enhanced Magnetic Resonance Imaging) series on small animals. In this approach we consider DCE-MRI series as multivariate images. A full multivariate segmentation method based on dimensionality reduction, noise filtering, supervised classification and stochastic watershed is explained and tested on several data sets. The two main key-points introduced in this paper are noise reduction preserving contours and spatio temporal segmentation by stochastic watershed. Noise reduction is performed in a special way that selects factorial axes of Factor Correspondence Analysis in order to preserves contours. Then a spatio-temporal approach based on stochastic watershed is used to segment tumours. The results obtained are in accordance with the diagnosis of the medical doctors.

Spatio-Colour Asplünd 's Metric and Logarithmic Image Processing for Colour Images (LIPC)

Asplund’s metric, which is useful for pattern matching, consists in a double-sided probing, i.e. the over-graph and the sub-graph of a function are probed jointly. This paper extends the Asplund’s metric we previously defined for colour and multivariate images using a marginal approach (i.e. component by component) to the first spatio-colour Asplund’s metric based on the vectorial colour LIP model (LIPC). LIPC is a non-linear model with operations between colour images which are consistent with the human visual system. The defined colour metric is insensitive to lighting variations and a variant which is robust to noise is used for colour pattern matching.

Contrast enhancement of eye fundus images

A significant number of digital eye fundus images have strong contrast variations which can be a limiting factor for the diagnosis of the diabetic retinopathy lesions. Currently, to address this problem, graders have to manually adjust the image contrast which is person dependent and therefore not easily reproducible. Images may still be considered un-gradable because they are too bright or too dark. We have developed a fully automatic method, which achieves contrast uniformity across the entire image. The method is based on a colour model consistent with the physical principles of image formation. The contrast of the dark or the bright elements are adjusted in a way that provides a similar colour aspect to lesions such as micro-aneurysms or to anatomical structures such as veins. This method is much more powerful than the previous existing grey level methods using polynomial adjustment, mathematical morphology or histogram equalisations. Our method has been tested on more than 2000 images acquired from different screening services ranging from a high resource country with quality controlled process while others were obtained from low resource countries under harsher conditions. Some images were bright while others were dark making diagnosis difficult. However for all images, the lighting variations have been corrected and the contrast has been enhanced for lesions such as micro-aneurysms and the vascular structures. They are now easier to be detected by graders. This new colour contrast method is a very promising tool to assist graders in diagnosing the presence of diabetic retinopathy and other lesions present in digital eye fundus images since the lesions appear to be much more evident in comparison of the original image. Importantly our method is fully automatic and can be easily integrated in a screening system.

Double-sided probing by map of Asplund's distances using Logarithmic Image Processing in the framework of Mathematical Morphology

We establish the link between Mathematical Morphology and the map of Asplunds distances between a probe and a grey scale function, using the Logarithmic Image Processing scalar multiplication. We demonstrate that the map is the logarithm of the ratio between a dilation and an erosion of the function by a structuring function: the probe. The dilations and erosions are mappings from the lattice of the images into the lattice of the positive functions. Using a flat structuring element, the expression of the map of Asplunds distances can be simplified with a dilation and an erosion of the image; these mappings stays in the lattice of the images. We illustrate our approach by an example of pattern matching with a non-flat structuring function.