Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

A comparison: Different DCNN models for intelligent object detection in remote sensing images

: Ding, P.; Zhang, Y.; Jia, P.; Chang, X.-L.


Neural processing letters 49 (2019), No.3, pp.1369-1379
ISSN: 1370-4621
ISSN: 1573-773X
Journal Article
Fraunhofer IGD ()

In recent years, deep learning especially deep convolutional neural networks (DCNN) has made great progress. Many researchers take advantage of different DCNN models to do object detection in remote sensing. Different DCNN models have different advantages and disadvantages. But in the field of remote sensing, many scholars usually do comparison between DCNN models and traditional machine learning. In this paper, we compare different state-of-the-art DCNN models mainly over two publicly available remote sensing datasets-airplane dataset and car dataset. Such comparison can provide guidance for related researchers. Besides,we provide suggestions for fine-tuning different DCNN models. Moreover, for DCNN models including fully connected layers, we provide a method to save storage space.