Benchmarking Deep Learning Frameworks for the Classification of Very High Resolution Satellite Multispectral Data

Papadomanolaki, Maria; Vakalopoulou, Maria; Zagoruyko, S.; Καράντζαλος, Κωνσταντίνος

doi:10.5194/isprsannals-iii-7-83-2016

Cited by 35 publications

(26 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After the evolution of AlexNet, which can be accepted as a milestone for deep learning, deeper architectures, such as visual geometry group (VGG) [17], GoogleNet [18], which came up with inception modules, and residual network (ResNet) [19], were developed and the error rate in the competition decreased gradually. Along with these advancements, researchers started to use CNN structures in object classification with satellite images [20][21][22][23][24]. Although the remote sensing images have less spatial details and complex background, these methods can achieve highly accurate results near the visual interpretation performance.…”

Section: Introductionmentioning

confidence: 99%

Comparative Research on Deep Learning Approaches for Airplane Detection from Very High-Resolution Satellite Images

2020

View full text Add to dashboard Cite

Object detection from satellite images has been a challenging problem for many years. With the development of effective deep learning algorithms and advancement in hardware systems, higher accuracies have been achieved in the detection of various objects from very high-resolution (VHR) satellite images. This article provides a comparative evaluation of the state-of-the-art convolutional neural network (CNN)-based object detection models, which are Faster R-CNN, Single Shot Multi-box Detector (SSD), and You Look Only Once-v3 (YOLO-v3), to cope with the limited number of labeled data and to automatically detect airplanes in VHR satellite images. Data augmentation with rotation, rescaling, and cropping was applied on the test images to artificially increase the number of training data from satellite images. Moreover, a non-maximum suppression algorithm (NMS) was introduced at the end of the SSD and YOLO-v3 flows to get rid of the multiple detection occurrences near each detected object in the overlapping areas. The trained networks were applied to five independent VHR test images that cover airports and their surroundings to evaluate their performance objectively. Accuracy assessment results of the test regions proved that Faster R-CNN architecture provided the highest accuracy according to the F1 scores, average precision (AP) metrics, and visual inspection of the results. The YOLO-v3 ranked as second, with a slightly lower performance but providing a balanced trade-off between accuracy and speed. The SSD provided the lowest detection performance, but it was better in object localization. The results were also evaluated in terms of the object size and detection accuracy manner, which proved that large-and medium-sized airplanes were detected with higher accuracy.

show abstract

Section: Introductionmentioning

confidence: 99%

Comparative Research on Deep Learning Approaches for Airplane Detection from Very High-Resolution Satellite Images

2020

View full text Add to dashboard Cite

show abstract

“…RF classifiers have been also reported to achieve high accuracy metrics in similar classification tasks [9,22,32], while studies have indicated that RF may present lower sensitivity to mislabelled training data than SVM [22]. On the other hand, deep learning is currently one of the fastest-growing trends in remote sensing data analysis, successfully tackling a variety of challenging problems and benchmark datasets in object recognition, semantic segmentation and image classification [33][34][35][36]. In particular, Convolutional Neural Networks (CNNs) are currently heavily used for classification and semantic segmentation tasks in a variety of remote sensing datasets [28,[37][38][39].…”

Section: Introductionmentioning

confidence: 99%

Detailed Land Cover Mapping from Multitemporal Landsat-8 Data of Different Cloud Cover

et al. 2018

View full text Add to dashboard Cite

Detailed, accurate and frequent land cover mapping is a prerequisite for several important geospatial applications and the fulfilment of current sustainable development goals. This paper introduces a methodology for the classification of annual high-resolution satellite data into several detailed land cover classes. In particular, a nomenclature with 27 different classes was introduced based on CORINE Land Cover (CLC) Level-3 categories and further analysing various crop types. Without employing cloud masks and/or interpolation procedures, we formed experimental datasets of Landsat-8 (L8) images with gradually increased cloud cover in order to assess the influence of cloud presence on the reference data and the resulting classification accuracy. The performance of shallow kernel-based and deep patch-based machine learning classification frameworks was evaluated. Quantitatively, the resulting overall accuracy rates differed within a range of less than 3%; however, maps produced based on Support Vector Machines (SVM) were more accurate across class boundaries and the respective framework was less computationally expensive compared to the applied patch-based deep Convolutional Neural Network (CNN). Further experimental results and analysis indicated that employing all multitemporal images with up to 30% cloud cover delivered relatively higher overall accuracy rates as well as the highest per-class accuracy rates. Moreover, by selecting 70% of the top-ranked features after applying a feature selection strategy, slightly higher accuracy rates were achieved. A detailed discussion of the quantitative and qualitative evaluation outcomes further elaborates on the performance of all considered classes and highlights different aspects of their spectral behaviour and separability.

show abstract

“…In particular, ConvNet has a relatively simple architecture. There are 4 blocks of layers: 2 convolutional and 2 fully-connected [56]. The first convolutional layer includes 3 × 3 kernels, a stride of 1 and padding equal to 0.…”

Section: Patch-based Learningmentioning

confidence: 99%

“…The AlexNet architecture is comprised of 8 blocks of layers following the same sequence: 5 convolutional and 3 fully-connected [56]. Giving some more details, the first convolutional layer applies 3 × 3 filters with a stride of 1, followed by a ReLU activation function and a max-pooling operation of kernels and stride equal to 3 × 3 and 2, respectively.…”

Section: Patch-based Learningmentioning

confidence: 99%

“…The VGG-16 architecture, which is much deeper [56] than the previous ones, consists of a block where the training data are subject to convolutional filters, batch normalization filters and a ReLU activation function. Data dimensions change only in terms of depth after being passed through this block since the convolution filters are of size 3 × 3 using a stride and padding of 1.…”

Section: Patch-based Learningmentioning

confidence: 99%

See 1 more Smart Citation

A Novel Object-Based Deep Learning Framework for Semantic Segmentation of Very High-Resolution Remote Sensing Data: Comparison with Convolutional and Fully Convolutional Networks

2019

View full text Add to dashboard Cite

Deep learning architectures have received much attention in recent years demonstrating state-of-the-art performance in several segmentation, classification and other computer vision tasks. Most of these deep networks are based on either convolutional or fully convolutional architectures. In this paper, we propose a novel object-based deep-learning framework for semantic segmentation in very high-resolution satellite data. In particular, we exploit object-based priors integrated into a fully convolutional neural network by incorporating an anisotropic diffusion data preprocessing step and an additional loss term during the training process. Under this constrained framework, the goal is to enforce pixels that belong to the same object to be classified at the same semantic category. We compared thoroughly the novel object-based framework with the currently dominating convolutional and fully convolutional deep networks. In particular, numerous experiments were conducted on the publicly available ISPRS WGII/4 benchmark datasets, namely Vaihingen and Potsdam, for validation and inter-comparison based on a variety of metrics. Quantitatively, experimental results indicate that, overall, the proposed object-based framework slightly outperformed the current state-of-the-art fully convolutional networks by more than 1% in terms of overall accuracy, while intersection over union results are improved for all semantic categories. Qualitatively, man-made classes with more strict geometry such as buildings were the ones that benefit most from our method, especially along object boundaries, highlighting the great potential of the developed approach.

show abstract

Benchmarking Deep Learning Frameworks for the Classification of Very High Resolution Satellite Multispectral Data

Cited by 35 publications

References 10 publications

Comparative Research on Deep Learning Approaches for Airplane Detection from Very High-Resolution Satellite Images

Comparative Research on Deep Learning Approaches for Airplane Detection from Very High-Resolution Satellite Images

Detailed Land Cover Mapping from Multitemporal Landsat-8 Data of Different Cloud Cover

A Novel Object-Based Deep Learning Framework for Semantic Segmentation of Very High-Resolution Remote Sensing Data: Comparison with Convolutional and Fully Convolutional Networks

Contact Info

Product

Resources

About