A multi level evaluation for fusion system interaction improvement
Résumé
Fusion systems for image interpretation are complex systems that involve a complete information treatment chain (from the information extraction to the decision). Local evaluation of all the sub-parts that composed the system is an interesting way to better characterize its behaviour but it generates many numerical indicators. This paper proposes two intermediate evaluations based on the construction of symbolic indicators from the numerical separability indexes. All the available quality information are then used into a progressive dashboard that allows to better interact with the system.