Sébastien Derivaux

Ontology-Based Object Recognition for Remote Sensing Image Interpretation

The multiplication of very high resolution (spatial or spectral) remote sensing images appears to be an opportunity to identify objects in urban and periurban areas. The classification methods applied in the object-oriented image analysis approach could be based on the use of domain knowledge. A major issue in these approaches is domain knowledge formalization and exploitation. In this paper, we propose a recognition method based on an ontology which has been developed by experts of the domain. In order to give objects a semantic meaning, we have developed a matching process between an object and the concepts of the ontology. Experiments are made on a Quickbird image. The quality of the results shows the effectiveness of the proposed method.

Supervised image segmentation using watershed transform, fuzzy classification and evolutionary computation

Automatic image interpretation is often achieved by first performing a segmentation of the image (i.e., gathering neighbouring pixels into homogeneous regions) and then applying a supervised region-based classification. In such a process, the quality of the segmentation step is of great importance in the final classified result. Nevertheless, whereas the classification step takes advantage from some prior knowledge such as learning sample pixels, the segmentation step rarely does. In this paper, we propose to involve such samples through machine learning procedures to improve the segmentation process. More precisely, we consider the watershed transform segmentation algorithm, and rely on both a fuzzy supervised classification procedure and a genetic algorithm in order to respectively build the elevation map used in the watershed paradigm and tune segmentation parameters. We also propose new criteria for segmentation evaluation based on learning samples. We have evaluated our method on remotely sensed images. The results assert the relevance of machine learning as a way to introduce knowledge within the watershed segmentation process.

Watershed Segmentation of Remotely Sensed Images Based on a Supervised Fuzzy Pixel Classification

Remotely sensed images are more and more precise (spatial resolution under 1 meter). For these images, objects of interest contains several pixels. Generally a segmentation method is used to cluster pixels that belong to the same objects before classification. The quality of such a segmentation method is crucial to achieve good clasification results. In this paper, a new segmentation method is proposed which aims to improve the classical watershed segmentation method based on multispectral gradient. The proposed method uses some labeled samples with classes of interest to induce a new dissimilarity between pixels which defines a new representation space to be used.

On Machine Learning in Watershed Segmentation

Automatic image interpretation could be achieved by first performing a segmentation of the image, i.e. aggregating similar pixels to form regions, then use a supervised region- based classification. In such a process, the quality of the segmentation step is of great importance. Nevertheless, whereas the classification step takes advantage from some prior knowledge such as learning sample pixels, the segmentation step rarely does. In this paper, we propose to involve machine learning to improve the segmentation process using the watershed transform. More precisely, we apply a fuzzy supervised classification and a genetic algorithm in order to respectively generate the elevation map used in the watershed transform and tune segmentation parameters. The results from our evolutionary supervised watershed algorithm confirm the relevance of machine learning to introduce knowledge in the watershed segmentation process.

Improving Supervised Learning with Multiple Clusterings

Classification task involves inducing a predictive model using a set of labeled samples. The accuracy of the model usually increases as more labeled samples are available. When one has only few samples, the obtained model tends to offer poor results. Even when labeled samples are difficult to get, a lot of unlabeled samples are generally available on which unsupervised learning can be done. In this chapter, a way to combine supervised and unsupervised learning in order to use both labeled and unlabeled samples is explored. The efficiency of the method is evaluated on various UCI datasets and on the classification of a very high resolution remote sensing image when the number of labeled samples is very low.

On the complementarity of an ontology and a nearest neighbour classifier for remotely sensed image interpretation

Automatic image interpretation could be achieved by first performing a segmentation of the image, i.e. aggregating similar pixels to form regions, then use a region-based classification. This paper presents two region-based classifications, namely a supervised classification and an ontology-based classification and discuss their pros and cons. As they are complementary, we propose to combine these two approaches. Results shown that the presented method is relevant.