Order Sensitive Scoring for Objective Evaluation of Remote Sensing - Image Retrieval System

R Kusumaningrum, M.I. Fanany, A.M. Arymurthy

Abstract


In order to make the evaluation result become fair and consistent without relying on user judgement, we proposed a new mechanism of objective evaluation for remote sensing image retrieval system. This mechanism is performed by dividing the image database into some subjects and select one subject as a ground truth. Subsequently, we take an image of the ground truth as a query Q. Retrieval accurate ratio (pAR) for query Q will be computed as a comparison between utility value and maximum possible utility. This pAR value can provide the non-degeneracy score. We repeat the retrieval process for all the images in the ground truth as the query and take averaged accurate ratio (pAAR) for a subject as the average value of all the accurate ratio for all queries in a subject. These process is performed for all subjects in the database, thus we can obtain the mean averaged accurate ratio (pmAAR) as mean value of all averaged accurate ratio for all subjects.

Full Text:

F17-F25

Refbacks

  • There are currently no refbacks.



Konferensi Nasional Teknologi Informasi & Aplikasinya
Jurusan Sistem Informasi | Fakultas Ilmu Komputer
Universitas Sriwijaya