Malignant melanoma is one of the most fatal forms of skin cancer. It has also become increasingly common, especially among white-skinned people exposed to the sun. Early detection of melanoma is essential to raise survival rates, since its detection at an early stage can be helpful and curable. Working out the dermoscopic clinical features (pigment network and lesion borders) of melanoma is a vital step for dermatologists, who require an accurate method of reaching the correct clinical diagnosis, and ensure the right area receives the correct treatment. These structures are considered one of the main keys that refer to melanoma or non-melanoma disease. However, determining these clinical features can be a time-consuming, subjective (even for trained clinicians) and challenging task for several reasons: lesions vary considerably in size and colour, low contrast between an affected area and the surrounding healthy skin, especially in early stages, and the presence of several elements such as hair, reflections, oils and air bubbles on almost all images. This thesis aims to provide an accurate, robust and reliable automated dermoscopy image analysis technique, to facilitate the early detection of malignant melanoma disease. In particular, four innovative methods are proposed for region segmentation and classification, including two for pigmented region segmentation, one for pigment network detection, and one for lesion classification. In terms of boundary delineation, four pre-processing operations, including Gabor filter, image sharpening, Sobel filter and image inpainting methods are integrated in the segmentation approach to delete unwanted objects (noise), and enhance the appearance of the lesion boundaries in the image. The lesion border segmentation is performed using two alternative approaches. The Fuzzy C-means and the Markov Random Field approaches detect the lesion boundary by repeating the labeling of pixels in all clusters, as a first method. Whereas, the Particle Swarm Optimization with the Markov Random Field method achieves greater accuracy for the same aim by combining them in the second method to perform a local search and reassign all image pixels to its cluster properly. With respect to the pigment network detection, the aforementioned pre-processing method is applied, in order to remove most of the hair while keeping the image information and increase the visibility of the pigment network structures. Therefore, a Gabor filter with connected component analysis are used to detect the pigment network lines, before several features are extracted and fed to the Artificial Neural Network as a classifier algorithm. In the lesion classification approach, the K-means is applied to the segmented lesion to separate it into homogeneous clusters, where important features are extracted; then, an Artificial Neural Network with Radial Basis Functions is trained by representative features to classify the given lesion as melanoma or not. The strong experimental results of the lesion border segmentation methods including Fuzzy C-means with Markov Random Field and the combination between the Particle Swarm Optimization and Markov Random Field, achieved an average accuracy of 94.00% , 94.74% respectively. Whereas, the lesion classification stage by using extracted features form pigment network structures and segmented lesions achieved an average accuracy of 90.1% , 95.97% respectively. The results for the entire experiment were obtained using a public database PH2 comprising 200 images. The results were then compared with existing methods in the literature, which have demonstrated that our proposed approach is accurate, robust, and efficient in the segmentation of the lesion boundary, in addition to its classification.

No description available.

No description available.

No description available.

In this thesis, a Content Based Image Retrieval (CBIR) system to query the objects in an image database is proposed. Images are represented as collections of regions after being segmented with Normalized Cuts algorithm. MPEG-7 content descriptors are used to encode regions in a 239-dimensional feature space. User of the proposed CBIR system decides which objects to query and labels exemplar regions to train the system using a graphical interface. Fuzzy ARTMAP algorithm is used to learn the mapping between feature vectors and binary coded class identification numbers. Preliminary recognition experiments prove the power of fuzzy ARTMAP as a region classifier. After training, features of all regions in the database are extracted and classified. Simple index files enabling fast access to all regions from a given class are prepared to be used in the querying phase. To retrieve images containing a particular object, user opens an image and selects a query region together with a label in the graphical interface of our system. Then the system ranks all regions in the indexed set of the query class with respect to their L2 (Euclidean) distance to the query region and displays resulting images. During retrieval experiments, comparable class precisions with respect to exhaustive searching of the database are maintained which demonstrates e ectiveness of the classifier in narrowing down the search space.

A soil climate parameterization is designed for peatland environments in the Canadian Land Surface Scheme (CLASS). Three wetland soil classes account for the variation in the hydraulic characteristics of organic soils. Saturated hydraulic conductivity varies from a median of 1.0 x 10-7 m/s in deeply humidified sapric peat to 2.8 x 10-4 m/s in relatively undecomposed fibric peat. Average pore volume fraction ranges from 0.83 to 0.93. Parameters are derived for the soil moisture characteristic curves of fibric, hemic and sapric peat, using the Campbell (1974) equation employed in CLASS, and the van Genuchten (1980) formulation. Validation of modelled water table depth and peat temperature is performed for a fen in northern Quebec and a bog in north-central Minnesota. The new parameterization results in more realistic simulation than the previous version of CLASS, which was constrained to using mineral soil properties to approximate those of organic soils. / Two approaches are used to model methane emissions from northern peatlands using the new soil climate parameterization in CLASS. In the first module, the multiple regression equation of Dise et al. (1993) is used to simulate daily methane emissions from water table depth and peat temperature. In the process-based module, methane flux is divided into its component parts: plant transport, diffusion and ebullition. Each of these transport mechanisms is determined by methane concentrations, which are calculated from a series of processes related to peat temperature, water table level and rooting depth. The daily methane emissions predicted by the two models are similar and correlate reasonably with observations from a bog in north-central Minnesota.

No description available.

This thesis examines the development and validation of Canadian Land Surface Scheme (CLASS) for various wetland landscapes individually, along with an evaluation of modelled results over a heterogeneous surface with airborne observations. A further statistical analysis of the effects of land surface classification procedures over the study area and their influence on modelled results is performed. CLASS is tested over individual wetland types: bog, fen and marsh in a stand-alone (non-GCM coupled) mode. Atmospheric conditions are provided for the eight site locations from tower measured data, while each surface is parameterized within the model from site specific measurements. Resulting model turbulent and radiative flux output is then statistically evaluated against observed tower data. Findings show that while CLASS models vascular dominated wetland areas (fen and marsh) quite well, non-vascular wetlands (bogs) are poorly represented, even with improved soil descriptions. At times when the water table is close to the surface, evaporation is greatly overestimated, whereas lowered water tables generate a vastly underestimated latent heat flux. Because CLASS does not include a moisture transfer scheme applicable for non-vascular vegetation, the description of this vegetation type as either a vascular plant or bare soil appears inappropriate. / CLASS was then tuned for a specific bog location found in the Hudson Bay Lowland (HBL) during the Northern Wetlands Study (NOWES). With bog surfaces better described within the model, testing of CLASS over a highly heterogeneous 169 km2 HBL region is then undertaken. The model is first modified for lake and pond surfaces and then separate runs for bog, fen, lake and tree/shrub categories is undertaken. Using a GIS, the test region under which airborne flux measurements are available is divided into 104 grid cells and proportions of each surface type are calculated within each cell. Findings indicate that although the modelled grid average radiation and flux values are reasonably well reproduced (4% error for net radiation, 10% for latent heat flux and 30% for sensible heat flux), spatial agreement between modelled and observed grid cells is disappointing. (Abstract shortened by UMI.)

We use a Fully Convolutional Neural Network (FCNN) to classify pixels in historical document images, enabling the extraction of high-quality, pixel-precise and semantically consistent layers of masked content. We also analyze a dataset of hand-labeled historical form images of unprecedented detail and complexity. The semantic categories we consider in this new dataset include handwriting, machine-printed text, dotted and solid lines, and stamps. Segmentation of document images into distinct layers allows handwriting, machine print, and other content to be processed and recognized discriminatively, and therefore more intelligently than might be possible with content-unaware methods. We show that an efficient FCNN with relatively few parameters can accurately segment documents having similar textural content when trained on a single representative pixel-labeled document image, even when layouts differ significantly. In contrast to the overwhelming majority of existing semantic segmentation approaches, we allow multiple labels to be predicted per pixel location, which allows for direct prediction and reconstruction of overlapped content. We perform an analysis of prevalent pixel-wise performance measures, and show that several popular performance measures can be manipulated adversarially, yielding arbitrarily high measures based on the type of bias used to generate the ground-truth. We propose a solution to the gaming problem by comparing absolute performance to an estimated human level of performance. We also present results on a recent international competition requiring the automatic annotation of billions of pixels, in which our method took first place.

We use a Fully Convolutional Neural Network (FCNN) to classify pixels in historical document images, enabling the extraction of high-quality, pixel-precise and semantically consistent layers of masked content. We also analyze a dataset of hand-labeled historical form images of unprecedented detail and complexity. The semantic categories we consider in this new dataset include handwriting, machine-printed text, dotted and solid lines, and stamps. Segmentation of document images into distinct layers allows handwriting, machine print, and other content to be processed and recognized discriminatively, and therefore more intelligently than might be possible with content-unaware methods. We show that an efficient FCNN with relatively few parameters can accurately segment documents having similar textural content when trained on a single representative pixel-labeled document image, even when layouts differ significantly. In contrast to the overwhelming majority of existing semantic segmentation approaches, we allow multiple labels to be predicted per pixel location, which allows for direct prediction and reconstruction of overlapped content. We perform an analysis of prevalent pixel-wise performance measures, and show that several popular performance measures can be manipulated adversarially, yielding arbitrarily high measures based on the type of bias used to generate the ground-truth. We propose a solution to the gaming problem by comparing absolute performance to an estimated human level of performance. We also present results on a recent international competition requiring the automatic annotation of billions of pixels, in which our method took first place.