Journal
REMOTE SENSING
Volume 10, Issue 6, Pages -Publisher
MDPI
DOI: 10.3390/rs10060964
Keywords
remote sensing image retrieval (RSIR); multi-label benchmark dataset; multi-label image retrieval; single-label image retrieval; handcrafted features; convolutional neural networks
Categories
Funding
- National key research and development plan on a Strategic International Scientific and Technological Innovation Cooperation Special Project [2016YFE0202300]
- Wuhan Chen Guang Project [2016070204010114]
- Guangzhou Science and Technology Project [201604020070]
- Special Task of Technical Innovation in Hubei Province [2016AAA018]
- Natural Science Foundation of China [61671332, 41771452, 41771454]
Ask authors/readers for more resources
Benchmark datasets are essential for developing and evaluating remote sensing image retrieval (RSIR) approaches. However, most of the existing datasets are single-labeled, with each image in these datasets being annotated by a single label representing the most significant semantic content of the image. This is sufficient for simple problems, such as distinguishing between a building and a beach, but multiple labels are required for more complex problems, such as RSIR. This motivated us to present a new benchmark dataset termed MLRSIR that was labeled from an existing single-labeled remote sensing archive. MLRSIR contained a total of 17 classes, and each image had at least one of 17 pre-defined labels. We evaluated the performance of RSIR methods ranging from traditional handcrafted feature-based methods to deep-learning-based ones on MLRSIR. More specifically, we compared the performances of RSIR methods from both single-label and multi-label perspectives. These results presented the advantages of multiple labels over single labels for interpreting complex remote sensing images, and serve as a baseline for future research on multi-label RSIR.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available