scholarly journals Benchmarking Anchor-Based and Anchor-Free State-of-the-Art Deep Learning Methods for Individual Tree Detection in RGB High-Resolution Images

2021 ◽  
Vol 13 (13) ◽  
pp. 2482
Author(s):  
Pedro Zamboni ◽  
José Marcato Junior ◽  
Jonathan de Andrade Silva ◽  
Gabriela Takahashi Miyoshi ◽  
Edson Takashi Matsubara ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Deep Learning ◽  
High Resolution ◽  
State Of The Art ◽  
Free State ◽  
Two Stage ◽  
Learning Methods ◽  
Single Tree ◽  
Sensing Applications ◽  
Remote Sensing Applications

Urban forests contribute to maintaining livability and increase the resilience of cities in the face of population growth and climate change. Information about the geographical distribution of individual trees is essential for the proper management of these systems. RGB high-resolution aerial images have emerged as a cheap and efficient source of data, although detecting and mapping single trees in an urban environment is a challenging task. Thus, we propose the evaluation of novel methods for single tree crown detection, as most of these methods have not been investigated in remote sensing applications. A total of 21 methods were investigated, including anchor-based (one and two-stage) and anchor-free state-of-the-art deep-learning methods. We used two orthoimages divided into 220 non-overlapping patches of 512 × 512 pixels with a ground sample distance (GSD) of 10 cm. The orthoimages were manually annotated, and 3382 single tree crowns were identified as the ground-truth. Our findings show that the anchor-free detectors achieved the best average performance with an AP50 of 0.686. We observed that the two-stage anchor-based and anchor-free methods showed better performance for this task, emphasizing the FSAF, Double Heads, CARAFE, ATSS, and FoveaBox models. RetinaNet, which is currently commonly applied in remote sensing, did not show satisfactory performance, and Faster R-CNN had lower results than the best methods but with no statistically significant difference. Our findings contribute to a better understanding of the performance of novel deep-learning methods in remote sensing applications and could be used as an indicator of the most suitable methods in such applications.

RNN-based multispectral satellite image processing for remote sensing applications

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Venkata Dasu Marri ◽  
Veera Narayana Reddy P. ◽  
Chandra Mohan Reddy S.
Keyword(s):  
Remote Sensing ◽  
Image Processing ◽  
Deep Learning ◽  
Image Classification ◽  
Satellite Image ◽  
Multispectral Image ◽  
Content Type ◽  
Colliding Bodies Optimization ◽  
Sensing Applications ◽  
Remote Sensing Applications

Purpose Image classification is a fundamental form of digital image processing in which pixels are labeled into one of the object classes present in the image. Multispectral image classification is a challenging task due to complexities associated with the images captured by satellites. Accurate image classification is highly essential in remote sensing applications. However, existing machine learning and deep learning–based classification methods could not provide desired accuracy. The purpose of this paper is to classify the objects in the satellite image with greater accuracy. Design/methodology/approach This paper proposes a deep learning-based automated method for classifying multispectral images. The central issue of this work is that data sets collected from public databases are first divided into a number of patches and their features are extracted. The features extracted from patches are then concatenated before a classification method is used to classify the objects in the image. Findings The performance of proposed modified velocity-based colliding bodies optimization method is compared with existing methods in terms of type-1 measures such as sensitivity, specificity, accuracy, net present value, F1 Score and Matthews correlation coefficient and type 2 measures such as false discovery rate and false positive rate. The statistical results obtained from the proposed method show better performance than existing methods. Originality/value In this work, multispectral image classification accuracy is improved with an optimization algorithm called modified velocity-based colliding bodies optimization.

scholarly journals Deep learning para la clasificación de usos de suelo agrícola con Sentinel-2

2020 ◽  
pp. 35
Author(s):  
M. Campos-Taberner ◽  
F.J. García-Haro ◽  
B. Martínez ◽  
M.A. Gilabert
Keyword(s):  
Remote Sensing ◽  
Deep Learning ◽  
Near Infrared ◽  
Spatial Information ◽  
Learning Algorithm ◽  
Support Vector ◽  
Sensing Applications ◽  
Learning Techniques ◽  
Remote Sensing Applications ◽  
Sentinel 2

<p class="p1">The use of deep learning techniques for remote sensing applications has recently increased. These algorithms have proven to be successful in estimation of parameters and classification of images. However, little effort has been made to make them understandable, leading to their implementation as “black boxes”. This work aims to evaluate the performance and clarify the operation of a deep learning algorithm, based on a bi-directional recurrent network of long short-term memory (2-BiLSTM). The land use classification in the Valencian Community based on Sentinel-2 image time series in the framework of the common agricultural policy (CAP) is used as an example. It is verified that the accuracy of the deep learning techniques is superior (98.6 % overall success) to that other algorithms such as decision trees (DT), k-nearest neighbors (k-NN), neural networks (NN), support vector machines (SVM) and random forests (RF). The performance of the classifier has been studied as a function of time and of the predictors used. It is concluded that, in the study area, the most relevant information used by the network in the classification are the images corresponding to summer and the spectral and spatial information derived from the red and near infrared bands. These results open the door to new studies in the field of the explainable deep learning in remote sensing applications.</p>

Detecting a keystone species European aspen in boreal forests with airborne hyperspectral, LiDAR and UAV data with machine learning methods

2021 ◽  
Author(s):  
Timo Kumpula ◽  
Janne Mäyrä ◽  
Anton Kuzmin ◽  
Arto Viinikka ◽  
Sonja Kivinen ◽  
...  
Keyword(s):  
Machine Learning ◽  
Remote Sensing ◽  
Deep Learning ◽  
High Resolution ◽  
Boreal Forests ◽  
Tree Level ◽  
Support Vector ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
European Aspen

&lt;p&gt;Sustainable forest management increasingly highlights the maintenance of biological diversity and requires up-to-date information on the occurrence and distribution of key ecological features in forest environments. Different proxy variables indicating species richness and quality of the sites are essential for efficient detecting and monitoring forest biodiversity. European aspen (Populus tremula L.) is a minor deciduous tree species with a high importance in maintaining biodiversity in boreal forests. Large aspen trees host hundreds of species, many of them classified as threatened. However, accurate fine-scale spatial data on aspen occurrence remains scarce and incomprehensive.&lt;/p&gt;&lt;p&gt;&amp;#160;&lt;/p&gt;&lt;p&gt;We studied detection of aspen using different remote sensing techniques in Evo, southern Finland. Our study area of 83 km&lt;sup&gt;2&lt;/sup&gt; contains both managed and protected southern boreal forests characterized by Scots pine (Pinus sylvestris L.), Norway spruce (Picea abies (L.) Karst), and birch (Betula pendula and pubescens L.), whereas European aspen has a relatively sparse and scattered occurrence in the area. We collected high-resolution airborne hyperspectral and airborne laser scanning data covering the whole study area and ultra-high resolution unmanned aerial vehicle (UAV) data with RGB and multispectral sensors from selected parts of the area. We tested the discrimination of aspen from other species at tree level using different machine learning methods (Support Vector Machines, Random Forest, Gradient Boosting Machine) and deep learning methods (3D convolutional neural networks).&lt;/p&gt;&lt;p&gt;&amp;#160;&lt;/p&gt;&lt;p&gt;Airborne hyperspectral and lidar data gave excellent results with machine learning and deep learning classification methods The highest classification accuracies for aspen varied between 91-92% (F1-score). The most important wavelengths for discriminating aspen from other species included reflectance bands of red edge range (724&amp;#8211;727 nm) and shortwave infrared (1520&amp;#8211;1564 nm and 1684&amp;#8211;1706 nm) (Viinikka et al. 2020; M&amp;#228;yr&amp;#228; et al 2021). Aspen detection using RGB and multispectral data also gave good results (highest F1-score of aspen = 87%) (Kuzmin et al 2021). Different remote sensing data enabled production of a spatially explicit map of aspen occurrence in the study area. Information on aspen occurrence and abundance can significantly contribute to biodiversity management and conservation efforts in boreal forests. Our results can be further utilized in upscaling efforts aiming at aspen detection over larger geographical areas using satellite images.&lt;/p&gt;

scholarly journals Deep learning in remote sensing applications: A meta-analysis and review

2019 ◽  
Vol 152 ◽  
pp. 166-177 ◽  
Author(s):  
Lei Ma ◽  
Yu Liu ◽  
Xueliang Zhang ◽  
Yuanxin Ye ◽  
Gaofei Yin ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Deep Learning ◽  
Meta Analysis ◽  
Sensing Applications ◽  
Remote Sensing Applications

scholarly journals Portable snapshot spectral imaging for agriculture

2018 ◽  
pp. 221-225 ◽  
Author(s):  
András Jung ◽  
René Michels ◽  
Rainer Graser
Keyword(s):  
Remote Sensing ◽  
High Resolution ◽  
Hyperspectral Imaging ◽  
Management Strategies ◽  
Variable Rate ◽  
Sensing Applications ◽  
Regular Variable ◽  
Remote Sensing Applications ◽  
Video Sensors ◽  
Areas Of Interest

High-resolution proximal and remote sensing applications can consolidate sustainable, prevention- and precision-oriented crop management strategies to decrease production risks. This paper shows significant perspectives, own developments and technical aspects of high resolution remote sensing in the context of field applications. Moreover, we provide an overview of snapshot hyperspectral imaging and potential field video sensors to identify areas of interest for their future development. One of the main conclusions of our paper is that non-scanning snapshot hyperspectral imaging technology may enable researchers to overcome the gap in the “point to image” transition of field sensing, while providing a flexible solution for regular variable-rate applications.

scholarly journals REMOTE SENSING APPLICATIONS IN THE FRAME OF “URBAN GEOLOGY” PROJECT

2017 ◽  
Vol 43 (3) ◽  
pp. 1486
Author(s):  
K. Nikolakopoulos ◽  
P. Tsompos
Keyword(s):  
Remote Sensing ◽  
Land Use ◽  
Image Analysis ◽  
High Resolution ◽  
Land Use Change ◽  
Satellite Data ◽  
Urban Geology ◽  
Sensing Applications ◽  
Remote Sensing Applications ◽  
Very High

In the frame of the “Urban Geology” project of IGME a lot of remote sensing applications were carried out: DSMs creation and accuracy verification, orthorectification of very high resolution satellite data, data fusion, multitemporal and multisensor image analysis, land cover and land use change detection e.t.c. The applications that took place in the pilot case of Nafplio are presented in this study

Unsupervised Representation High-Resolution Remote Sensing Image Scene Classification via Contrastive Learning Convolutional Neural Network

2021 ◽  
Vol 87 (8) ◽  
pp. 577-591
Author(s):  
Fengpeng Li ◽  
Jiabao Li ◽  
Wei Han ◽  
Ruyi Feng ◽  
Lizhe Wang
Keyword(s):  
Neural Network ◽  
Remote Sensing ◽  
Deep Learning ◽  
High Resolution ◽  
Convolutional Neural Network ◽  
State Of The Art ◽  
Remote Sensing Image ◽  
Scene Classification ◽  
Data Set ◽  
Unsupervised Deep Learning

Inspired by the outstanding achievement of deep learning, supervised deep learning representation methods for high-spatial-resolution remote sensing image scene classification obtained state-of-the-art performance. However, supervised deep learning representation methods need a considerable amount of labeled data to capture class-specific features, limiting the application of deep learning-based methods while there are a few labeled training samples. An unsupervised deep learning representation, high-resolution remote sensing image scene classification method is proposed in this work to address this issue. The proposed method, called contrastive learning, narrows the distance between positive views: color channels belonging to the same images widens the gaps between negative view pairs consisting of color channels from different images to obtain class-specific data representations of the input data without any supervised information. The classifier uses extracted features by the convolutional neural network (CNN)-based feature extractor with labeled information of training data to set space of each category and then, using linear regression, makes predictions in the testing procedure. Comparing with existing unsupervised deep learning representation high-resolution remote sensing image scene classification methods, contrastive learning CNN achieves state-of-the-art performance on three different scale benchmark data sets: small scale RSSCN7 data set, midscale aerial image data set, and large-scale NWPU-RESISC45 data set.

scholarly journals High-resolution, non-scanning FTS for remote sensing

1995 ◽  
Vol 149 ◽  
pp. 338-339
Author(s):  
K. Døhlen ◽  
A. Cañas
Keyword(s):  
Remote Sensing ◽  
Fourier Transform ◽  
High Resolution ◽  
Near Infrared ◽  
Michelson Interferometer ◽  
Resolving Power ◽  
Diode Array ◽  
Sensing Applications ◽  
First Results ◽  
Remote Sensing Applications

We present the first results from a portable spectrometer for the visible and very near infrared based upon the principle of heterodyned holographic Fourier transform spectroscopy (HHS) (Dohi and Suzuki 1971, Dohlen 1994). The instrument uses a Michelson interferometer where one of the mirrors is replaced with a grating. This produces a spatially located, frequency-shifted interferogram which is read out by an all-reflective relay lens and a photo-diode array and processed on a portable PC. A battery pack ensures an autonomy of about 7 hours. Instrumental assets include high optical throughput, variable resolving power, and no moving parts.We have successfully used the instrument in two different remote sensing applications: detection of vegetation reflectance and atmospheric absorption.

scholarly journals DOMAIN ADAPTATION WITH CYCLEGAN FOR CHANGE DETECTION IN THE AMAZON FOREST

2020 ◽  
Vol XLIII-B3-2020 ◽  
pp. 1635-1643
Author(s):  
P. J. Soto ◽  
G. A. O. P. Costa ◽  
R. Q. Feitosa ◽  
P. N. Happ ◽  
M. X. Ortega ◽  
...  
Keyword(s):  
Remote Sensing ◽  
Deep Learning ◽  
Change Detection ◽  
Domain Adaptation ◽  
Fine Tuning ◽  
Classification Model ◽  
Generative Adversarial Networks ◽  
Amazon Forest ◽  
Sensing Applications ◽  
Remote Sensing Applications

Abstract. Deep learning classification models require large amounts of labeled training data to perform properly, but the production of reference data for most Earth observation applications is a labor intensive, costly process. In that sense, transfer learning is an option to mitigate the demand for labeled data. In many remote sensing applications, however, the accuracy of a deep learning-based classification model trained with a specific dataset drops significantly when it is tested on a different dataset, even after fine-tuning. In general, this behavior can be credited to the domain shift phenomenon. In remote sensing applications, domain shift can be associated with changes in the environmental conditions during the acquisition of new data, variations of objects’ appearances, geographical variability and different sensor properties, among other aspects. In recent years, deep learning-based domain adaptation techniques have been used to alleviate the domain shift problem. Recent improvements in domain adaptation technology rely on techniques based on Generative Adversarial Networks (GANs), such as the Cycle-Consistent Generative Adversarial Network (CycleGAN), which adapts images across different domains by learning nonlinear mapping functions between the domains. In this work, we exploit the CycleGAN approach for domain adaptation in a particular change detection application, namely, deforestation detection in the Amazon forest. Experimental results indicate that the proposed approach is capable of alleviating the effects associated with domain shift in the context of the target application.

Sign in / Sign up

Export Citation Format

Share Document