Abstract | ||
---|---|---|
This paper describes a fast multilabel classification method for unmanned aerial vehicle (UAV) images acquired over urban areas. It starts by subdividing a given query image into a set of equal tiles, which are successively processed and analyzed separately. In particular, each tile is described by extracting opportune features which are then further transformed through the learning of an Autoencoder (AE) model. This last provides new features of reduced dimensionality, exploited to feed a multilayer perceptron (MLP) classifier in order to derive the list of objects present in the considered tile. From the conducted experiments, it comes out that the proposed method yields interesting classification accuracies and much shorter processing times compared to the state-of-the-art. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1109/JURSE.2017.7924544 | 2017 Joint Urban Remote Sensing Event (JURSE) |
Keywords | Field | DocType |
multilabeling UAV images,unmanned aerial vehicle,autoencoder networks,multilabel classification method,urban areas,query image,autoencoder model,reduced dimensionality features,multilayer perceptron classifier | Computer vision,Autoencoder,Computer science,Curse of dimensionality,Multilayer perceptron,Artificial intelligence,Classifier (linguistics),Tile | Conference |
ISSN | ISBN | Citations |
2334-0932 | 978-1-5090-5809-9 | 0 |
PageRank | References | Authors |
0.34 | 10 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Abdallah Zeggada | 1 | 43 | 4.12 |
Farid Melgani | 2 | 1100 | 80.98 |