Analyse du parcellaire par traitement de données multitemporelles S2 à l’aide d’une méthode de fouille de données = Ứng dụng phương pháp khai phá dữ liệu cho phân tích lô bằng xử lý dữ liệu đa thời gian S2.

• Main Author: Rodrique, KAFANDO
• Other Authors: Ienco, Dino
• Format: Theses
• Language: French
• Published: 2020
• Subjects: Công nghệ thông tin ; Khai phá dữ liệu ; Cơ sở dữ liệu; 005.74
• Online Access: http://repository.vnu.edu.vn/handle/VNU_123/69295
• Institution: Vietnam National University, Hanoi

Food risk managment is nowadays one of the scourges that threatens the world, especially for developing countries. Faced to this situation, it becomes necessary to take some measures that will help to control the various farms according to their occupations and their changes during the seasons. Agricultural occupations are difficult to control, especially in areas with limited access. An occupation can vary from a period to another, either on its morphological and/or its content. The main question then is how to automatically monitor the behaviour of a given agricultural parcel over an STIS , regardless of its size. It’s within this study that a time series analysis methodology has been set up by [4] in order to be able to automatically monitor the evolution of a given entity considering the morphological level. Thanks to this new approach, it is now possible to monitor the evolution of an agricultural plot on which we have satellite image time series and to carry out analyses on this plot. In this study, we will study more precisely the stability of the cultural plot of the Kamech basin, an agricultural zone located in the southern of Tunisia. Carried out within the UMR-TETIS research unit, precisely at IRSTEA over a period of six (06) months, this report has two main aims. The first is an academic requierement that allow the qualification of our Master’s degree program in computer science. The second is a professional one, because our work was developed as part of a research and innovation project