Markov, graphical models and statistical decision in image: From statistical formalization to technology transfer

The author briefly recalls his academic background and his previous work on image research using statistical approaches (Stochastic modeling and exploitation of Markovian fields). The motivation for this work came from his practice of Markovian fields in applications during the first decade of his research career. It is clear that the learning difficulties can limit their interest despite the ease with which it is possible to integrate physics and application objectives of the images to be analyzed. In order to appreciate the contributions in the field of statistics and stochastic modelling, she synthesizes here this approach in three chapters, each one focused on one of the three main tools at the heart of her contributions: Markovian fields, graphical models and statistical decision. In each chapter, the central contribution and the main results are presented. The scientific production is of course recalled as well as the supervisions and the publications. Each chapter ends with a conclusion reporting her vision and perspectives that she would be tempted to implement on the tool as well as on the applications to which the tool has contributed.