scholarly journals Mapping Forest Types in China with 10 m Resolution Based on Spectral–Spatial–Temporal Features

2021 ◽  
Vol 13 (5) ◽  
pp. 973
Author(s):  
Kai Cheng ◽  
Juanle Wang ◽  
Xinrong Yan
Keyword(s):  
Remote Sensing ◽  
Forest Type ◽  
Google Earth ◽  
Growth Patterns ◽  
Forest Types ◽  
Forest Classification ◽  
Classification Framework ◽  
Temporal Features ◽  
Occurrence Matrix ◽  
Type Classification

The comprehensive application of spectral, spatial, and temporal (SST) features derived from remote sensing images is a significant technique for classifying and mapping forest types. Facing limitations in the availability of detailed forest type identification processes for large regions, a forest type classification framework based on SST features was developed in this study. The advantages of Sentinel-2 and Landsat series imagery were used to extract SST forest type classification features, using red-edge bands, a gray-level co-occurrence matrix, and harmonic analysis, with the assistance of the Google Earth Engine platform. Considering four representative Chinese geographic regions—middle and high latitudes, complex mountainous areas, cloudy and rainy areas, and the N–S climate transition zone—our method was proven to be effective, with overall classification accuracies > 85%. The scheme to assess the importance of SST features for forest classification in various regions was designed using the Gini criterion in the random forest algorithm and revealed that spectral features were more effective in classifying forest types with complex compositions. Temporal features were found to be favorable for identifying forest types with obvious evergreen and deciduous growth patterns, while spatial features produced better classification results for forest types with different spatial structures, such as needle- or broad-leaved forests. The findings of this study can provide a reference for feature selection in remote sensing forest type classification processes, and identifying forest types in this way could provide support for the accurate and sustainable management of forest resources.

scholarly journals Potential of P-Band SAR Tomography in Forest Type Classification

2021 ◽  
Vol 13 (4) ◽  
pp. 696
Author(s):  
Dinh Ho Tong Minh ◽  
Yen-Nhi Ngo ◽  
Thu Trang Lê
Keyword(s):  
Forest Type ◽  
Three Dimensional ◽  
Strong Support ◽  
Low Frequency ◽  
High Sensitivity ◽  
Forest Site ◽  
Forest Types ◽  
Forest Classification ◽  
Sar Data ◽  
Type Classification

Forest type classification using spaceborne remote sensing is a challenge. Low-frequency Synthetic Aperture Radar (SAR) signals (i.e., P-band, ∼0.69 m wavelength) are needed to penetrate a thick vegetation layer. However, this measurement alone does not guarantee a good performance in forest classification tasks. SAR tomography, a technique employing multiple acquisitions over the same areas to form a three-dimensional image, has been demonstrated to improve SAR’s capability in many applications. Our study shows the potential value of SAR tomography acquisitions to improve forest classification. By using P-band tomographic SAR data from the German Aerospace Center F-SAR sensor during the AfriSAR campaign in February 2016, the vertical profiles of five different forest types at a tropical forest site in Mondah, Gabon (South Africa) were analyzed and exploited for the classification task. We demonstrated that the high sensitivity of SAR tomography to forest vertical structure enables the improvement of classification performance by up to 33%. Interestingly, by using the standard Random Forest technique, we found that the ground (i.e., at 5–10 m) and volume layers (i.e., 20–40 m) play an important role in identifying the forest type. Together, these results suggested the promise of the TomoSAR technique for mapping forest types with high accuracy in tropical areas and could provide strong support for the next Earth Explorer BIOMASS spaceborne mission which will collect P-band tomographic SAR data.

scholarly journals Self-Organizing Deep Learning (SO-UNet)—A Novel Framework to Classify Urban and Peri-Urban Forests

2021 ◽  
Vol 13 (10) ◽  
pp. 5548
Author(s):  
Mohamad M. Awad ◽  
Marco Lauteri
Keyword(s):  
Remote Sensing ◽  
Deep Learning ◽  
Spatial Resolution ◽  
Forest Type ◽  
Urban Forests ◽  
Learning Technologies ◽  
Aerial Images ◽  
Hyper Spectral ◽  
Type Classification ◽  
Self Organizing

Forest-type classification is a very complex and difficult subject. The complexity increases with urban and peri-urban forests because of the variety of features that exist in remote sensing images. The success of forest management that includes forest preservation depends strongly on the accuracy of forest-type classification. Several classification methods are used to map urban and peri-urban forests and to identify healthy and non-healthy ones. Some of these methods have shown success in the classification of forests where others failed. The successful methods used specific remote sensing data technology, such as hyper-spectral and very high spatial resolution (VHR) images. However, both VHR and hyper-spectral sensors are very expensive, and hyper-spectral sensors are not widely available on satellite platforms, unlike multi-spectral sensors. Moreover, aerial images are limited in use, very expensive, and hard to arrange and manage. To solve the aforementioned problems, an advanced method, self-organizing–deep learning (SO-UNet), was created to classify forests in the urban and peri-urban environment using multi-spectral, multi-temporal, and medium spatial resolution Sentinel-2 images. SO-UNet is a combination of two different machine learning technologies: artificial neural network unsupervised self-organizing maps and deep learning UNet. Many experiments have been conducted, and the results showed that SO-UNet overwhelms UNet significantly. The experiments encompassed different settings for the parameters that control the algorithms.

scholarly journals Russian Approaches to the Forest Type Classification

2021 ◽  
Vol 906 (1) ◽  
pp. 012023
Author(s):  
Valery Fomin ◽  
Anna Mikhailovich
Keyword(s):  
Geographical Distribution ◽  
Russian Federation ◽  
Forest Type ◽  
Anthropogenic Factors ◽  
Genetic Approach ◽  
Habitat Conditions ◽  
Forest Types ◽  
The Russian Federation ◽  
Forest Typology ◽  
Type Classification

Abstract The results of researches characterizing the geographical distribution of forest-ecological, phytocoenotic, and genetic classifications of forest types in the Russian Federation nowadays are presented in the thesis. A comparative analysis was carried out for the following items: the inclusive concept of a classification unit (a type of habitat conditions; a type of forest); features of distinguishing the border of the classification units; classification features used to determine the type of habitat conditions; features of the classification of phytocoenoses used to determine the forest type; the degree to which the successional dynamics of forest stands are taken into consideration; the degree to which the influence of anthropogenic factors are taken into consideration; the level of implementation in forest management and forestry practice; regions of implementation. In the process of development of forest typologies, the concept of a forest type changed from understanding it as a forest area homogeneous by composition, structure, and appearance (homogeneity in space) in natural classifications to the concepts of a forest type, in which priority is given to homogeneity in origin (genesis), as well as developmental processes and dynamics (homogeneity in time) in genetic and dynamic typologies. Currently, there is the following forest type classification in the Russian Federation: forest-ecological, phytocoenotic, genetic, and dynamic. When classifying forest areas within the forest-ecological direction provided by E.V. Alekseev – P.S. Pogrebnyak, the priority is given to the characteristics of the habitat conditions. Within the phytocoenotic direction provided by V.N. Sukachev, the priority is given to the phytocoenosis characteristics. Within the genetic approach provided by B.A. Ivashkevich – B.P. Kolesnikov, a forest type is considered as a series of alternating phases – types of phytocoenosis within the same type of habitat conditions. In this case, phytocoenotic classifications can be a part of the genetic classifications for the climax forest phytocoenosis. And the dynamic approach provided by I.S. Melekhov is very close to the genetic one and is a superstructure over the classical phytocoenotic forest typology provided by V.N. Sukachev. The current use of forest typological classifications by forest inventory management enterprises in the Russian Federation was studied. A map of the geographical distribution of forest typologies of the above-listed directions of forest typology researches was created. Forest-ecological classifications are used mainly in the southern regions of the European part of Russia and the North Caucasus. Forest typologies created based on a genetic approach to the forest type classification are used in Western Siberia, in the south of the Far East and Eastern Siberia, and in some regions of the Urals. Phytocoenotic classifications of forest types are used in other regions of the Russian Federation.

scholarly journals MONITORING AND ASSESSMENT OF AGRI-URBAN LAND CONVERSION USING MULTI-SENSOR REMOTE SENSING AND GIS TECHNIQUES

2021 ◽  
Vol V-3-2021 ◽  
pp. 117-124
Author(s):  
D. C. Fargas Jr. ◽  
G. A. M. Narciso ◽  
A. C. Blanco
Keyword(s):  
Remote Sensing ◽  
Agricultural Land ◽  
Google Earth ◽  
High Rate ◽  
Landsat 8 ◽  
Land Conversion ◽  
Lulc Change ◽  
Random Forest Classification ◽  
Forest Classification ◽  
Single Sensor

Abstract. Continuous agricultural land conversion poses threat to food security but this has not been monitored due to ineffectual policies. One of the Philippine provinces with a high rate of conversion is the rice-producing province of Cavite. To assess the spatiotemporal dynamics of agricultural land conversion in Cavite, this study aims to develop an operational methodology to produce Land Use and Land Cover (LULC) change maps using a multi-sensor remote sensing approach for decision making and planning. LULC maps were generated using Random Forest Classification of Landsat 8 and Sentinel-1 image collections. Spectral indices, combinations of radar polarizations (VV, VH), and their principal components were included to improve its accuracy. Conversion maps were generated by taking the bi-annual difference of LULC maps from 2016 to 2019. Accuracy was assessed using visual inspection with Google Earth Pro. Classification was carried out using single-sensor (optical or radar) and multi-sensor (optical and radar) approach in combination with three feature selection algorithms, namely, Sandri and Zuccolotto (2006), Liaw and Wiener (2015), Kursa and Rudnicki (2010). Multi-sensor and single sensor yielded similarly high overall accuracies (OA = 96%) with the exception of single-sensor radar approach (OA = 53%). Multi-sensor approaches exhibit high accuracies (Cumulative Accuracy = 91%) in detecting agricultural to built-up LULC change up to 5,000 square meters unlike single-sensor optical approach (Cumulative Accuracy = 76%). Among the multi-sensor approaches, the method of Liaw and Wiener (2015) remains to be superior as it only uses eight (8) variables.

scholarly journals Multi-Label Remote Sensing Image Classification with Latent Semantic Dependencies

2020 ◽  
Vol 12 (7) ◽  
pp. 1110
Author(s):  
Junchao Ji ◽  
Weipeng Jing ◽  
Guangsheng Chen ◽  
Jingbo Lin ◽  
Houbing Song
Keyword(s):  
Remote Sensing ◽  
Image Classification ◽  
Short Term Memory ◽  
Satellite Image ◽  
Remote Sensing Image ◽  
Amazon Rainforest ◽  
Classification Framework ◽  
Remote Sensing Image Classification ◽  
Semantic Dependencies ◽  
Occurrence Matrix

Deforestation in the Amazon rainforest results in reduced biodiversity, habitat loss, climate change, and other destructive impacts. Hence obtaining location information on human activities is essential for scientists and governments working to protect the Amazon rainforest. We propose a novel remote sensing image classification framework that provides us with the key data needed to more effectively manage deforestation and its consequences. We introduce the attention module to separate the features which are extracted from CNN(Convolutional Neural Network) by channel, then further send the separated features to the LSTM(Long-Short Term Memory) network to predict labels sequentially. Moreover, we propose a loss function by calculating the co-occurrence matrix of all labels in the dataset and assigning different weights to each label. Experimental results on the satellite image dataset of the Amazon rainforest show that our model obtains a better F 2 score compared to other methods, which indicates that our model is effective in utilizing label dependencies to improve the performance of multi-label image classification.

scholarly journals Forest Type Classification Based on Integrated Spectral-Spatial-Temporal Features and Random Forest Algorithm—A Case Study in the Qinling Mountains

Forests ◽  
2019 ◽  
Vol 10 (7) ◽  
pp. 559 ◽  
Author(s):  
Kai Cheng ◽  
Juanle Wang
Keyword(s):  
Time Series ◽  
Random Forest ◽  
Forest Type ◽  
Evergreen Forest ◽  
Recursive Feature Elimination ◽  
Validation Dataset ◽  
Qinling Mountains ◽  
Temporal Features ◽  
Time Series Images ◽  
Type Classification

Spectral, spatial, and temporal features play important roles in land cover classification. However, limitations still exist in the integrated application of spectral-spatial-temporal (SST) features for forest type discrimination. This paper proposes a forest type classification framework based on SST features and the random forest (RF) algorithm. The SST features were derived from time-series images using original bands, vegetation index, gray-level correlation matrix, and harmonic analysis. Random forest-recursive feature elimination (RF-RFE) was used to optimize high-dimensional and correlated feature space, and determine the optimal SST feature set. Then, the classification was carried out using an RF classifier and the optimized SST feature set. This method was applied in the Qinling Mountains using Sentinel-2 time-series images. A total of 21 SST features were obtained through the RF-RFE method, and their importance was evaluated using the Gini index. The results indicated that spectral features contribute the most to separating shrubs, spatial features are more suitable for discrimination among evergreen forest types, and temporal features are more useful for evergreen forest, deciduous forest, and shrub types. The forest type map was generated based on the optimal SST feature set and RF algorithm, and evaluated based on an agreement with the validation dataset. The results showed that this integrated method is reliable, with an overall accuracy of 86.88% and kappa coefficient of 0.86, and can support forest type sustainable management and mapping at the local scale.

scholarly journals MAPPING SPATIAL ACCURACY OF FOREST TYPE CLASSIFICATION IN JAXA’s HIGH-RESOLUTION LAND USE AND LAND COVER MAP

2019 ◽  
Vol IV-3/W1 ◽  
pp. 57-63
Author(s):  
N. Tsutsumida ◽  
S. Nagai ◽  
P. Rodríguez-Veiga ◽  
J. Katagi ◽  
K. Nasahara ◽  
...  
Keyword(s):  
Land Use ◽  
Land Cover ◽  
Forest Type ◽  
Accuracy Assessment ◽  
Spatial Accuracy ◽  
Forest Types ◽  
Broadleaf Forest ◽  
Northern Edge ◽  
Accuracy Measures ◽  
Type Classification

<p><strong>Abstract.</strong> Accuracy assessment of forest type maps is essential to evaluate the classification of forest ecosystems quantitatively. However, map users do not understand in which regions those forest types are well classified from conventional static accuracy measures. Hence, the objective of this study is to unveil spatial heterogeneities of accuracies of forest type classification in a map. Four forest types (deciduous broadleaf forest (DBF), deciduous needleleaf forest (DNF), evergreen broadleaf forest (EBF), and evergreen needleleaf forest (ENF)) found in the JAXA’s land use / cover map of Japan were assessed by a volunteered Site-based dataset for Assessment of Changing LAnd cover by JAXA (SACLAJ). A geographically weighted (GW) correspondence matrix was applied to them to calculate the degree of overall agreements of forest type classes (forest overall accuracy), and the degree of accuracy for each forest class (forest user’s and producer’s accuracies) in a spatially varying way. This study compared spatial surfaces of these measures with static ones of them. The results show that the forest overall accuracy of the forest map tends to be relatively more accurate in the central Japan, while less in the Kansai and Chubu regions and the northern edge of Hokkaido. Static forest user’s accuracy measures for DBF, DNF, and ENF are better than forest producer’s accuracy ones, while the GW approach tells us such characteristics vary spatially and some areas have opposite trends. This kind of spatial accuracy assessment provides a more informative description of the accuracy than the simple use of conventional accuracy measures.</p>

scholarly journals Assessing the Effect of Training Sampling Design on the Performance of Machine Learning Classifiers for Land Cover Mapping Using Multi-Temporal Remote Sensing Data and Google Earth Engine

2021 ◽  
Vol 13 (8) ◽  
pp. 1433
Author(s):  
Shobitha Shetty ◽  
Prasun Kumar Gupta ◽  
Mariana Belgiu ◽  
S. K. Srivastav
Keyword(s):  
Machine Learning ◽  
Remote Sensing ◽  
Random Sampling ◽  
Sampling Design ◽  
Remote Sensing Data ◽  
Google Earth ◽  
Machine Learning Classifiers ◽  
Learning Classifiers ◽  
Multi Temporal ◽  
Google Earth Engine

Machine learning classifiers are being increasingly used nowadays for Land Use and Land Cover (LULC) mapping from remote sensing images. However, arriving at the right choice of classifier requires understanding the main factors influencing their performance. The present study investigated firstly the effect of training sampling design on the classification results obtained by Random Forest (RF) classifier and, secondly, it compared its performance with other machine learning classifiers for LULC mapping using multi-temporal satellite remote sensing data and the Google Earth Engine (GEE) platform. We evaluated the impact of three sampling methods, namely Stratified Equal Random Sampling (SRS(Eq)), Stratified Proportional Random Sampling (SRS(Prop)), and Stratified Systematic Sampling (SSS) upon the classification results obtained by the RF trained LULC model. Our results showed that the SRS(Prop) method favors major classes while achieving good overall accuracy. The SRS(Eq) method provides good class-level accuracies, even for minority classes, whereas the SSS method performs well for areas with large intra-class variability. Toward evaluating the performance of machine learning classifiers, RF outperformed Classification and Regression Trees (CART), Support Vector Machine (SVM), and Relevance Vector Machine (RVM) with a >95% confidence level. The performance of CART and SVM classifiers were found to be similar. RVM achieved good classification results with a limited number of training samples.

Sign in / Sign up

Export Citation Format

Share Document