An Artificial Neural Network Combined to Object Oriented Method for Land Cover Classification of High Resolution RGB Remote Sensing Images

Baroud, Sohaib; Chokri, Soumia; Belhaous, Safa; Hidila, Zineb; Mestari, Mohammed

doi:10.1007/978-3-030-45183-7_17

Cited by 2 publications

(2 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition Jin et al [27] presented a method that combines object-oriented approach with deep convolutional neural networks (COCNN). Baroud et al [28] also proposed an artificial neural network combined with object-oriented method for land cover classification of high-resolution multispectral remote sensing images.…”

Section: Object-oriented Classificationmentioning

confidence: 99%

Land Use Classification of High-Resolution Multispectral Satellite Images With Fine-Grained Multiscale Networks and Superpixel Postprocessing

Deng

Wei

2023

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Land use recognition from multispectral satellite images is fundamentally critical for geological applications, but the results are not satisfied. The scale dimension of current multiscale learning is too coarse to account for rich scales in multispectral images, and pixel-wise classification tends to produce "salt-and-pepper" labels due to possible misclassification in heterogeneous regions. In this paper, these issues are addressed by proposing a new pixel-wise classification model with finer scales for convolutional neural networks. The model is designed to extract multiscale contextual information using multiscale networks at a fine-grained level, addressing the issue of insufficient multiscale learning for classification. Furthermore, a smallscale segmentation-combination method is introduced as a postprocessing solution to smooth fragmented classification results. The proposed method is tested on GF-1, GF-2, DEIMOS-2, GeoEye-1, and Sentinel-2 satellite images, and compared with six neural-network-based algorithms. The results demonstrate the effectiveness of the proposed model in finding objects of large scale difference, improving classification accuracy, and reducing classified fragments. The discussion also illustrates that convolutional neural networks and pixel-wise inference are more practical than Transformer and patch-wise recognition.

show abstract

Section: Object-oriented Classificationmentioning

confidence: 99%

Land Use Classification of High-Resolution Multispectral Satellite Images With Fine-Grained Multiscale Networks and Superpixel Postprocessing

Deng

Wei

2023

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Many studies conducted land cover change detection and prediction using different models of fuzzy logic modeling [16][17][18][19][20]; geo-statistical methods [21][22][23][24]; Markov-CA [5,10,[25][26][27][28][29][30][31][32]; cellular automata models [33][34][35][36]; propagating aleatory and epistemic uncertainty [37,38]; artificial neural network [39][40][41][42][43]; Hopfield neural network [44][45][46][47][48]; supervised back-propagation neural network [49,50]; self-adaptive cellular based deep learning [51][52][53][54]; analytical hierarchy process [27,55,56]; geographic information system (GIS)-based hybrid site condition [15,57,58]; recurrent neural network…”

Section: Introductionmentioning

confidence: 99%

Extraction of Land Information, Future Landscape Changes and Seismic Hazard Assessment: A Case Study of Tabriz, Iran

Mohammadi

Karimzadeh

Kamran

et al. 2020

Sensors

View full text Add to dashboard Cite

Exact land cover inventory data should be extracted for future landscape prediction and seismic hazard assessment. This paper presents a comprehensive study towards the sustainable development of Tabriz City (NW Iran) including land cover change detection, future potential landscape, seismic hazard assessment and municipal performance evaluation. Landsat data using maximum likelihood (ML) and Markov chain algorithms were used to evaluate changes in land cover in the study area. The urbanization pattern taking place in the city was also studied via synthetic aperture radar (SAR) data of Sentinel-1 ground range detected (GRD) and single look complex (SLC). The age of buildings was extracted by using built-up areas of all classified maps. The logistic regression (LR) model was used for creating a seismic hazard assessment map. From the results, it can be concluded that the land cover (especially built-up areas) has seen considerable changes from 1989 to 2020. The overall accuracy (OA) values of the produced maps for the years 1989, 2005, 2011 and 2020 are 96%, 96%, 93% and 94%, respectively. The future potential landscape of the city showed that the land cover prediction by using the Markov chain model provided a promising finding. Four images of 1989, 2005, 2011 and 2020, were employed for built-up areas’ land information trends, from which it was indicated that most of the built-up areas had been constructed before 2011. The seismic hazard assessment map indicated that municipal zones of 1 and 9 were the least susceptible areas to an earthquake; conversely, municipal zones of 4, 6, 7 and 8 were located in the most susceptible regions to an earthquake in the future. More findings showed that municipal zones 1 and 4 demonstrated the best and worst performance among all zones, respectively.

show abstract

An Artificial Neural Network Combined to Object Oriented Method for Land Cover Classification of High Resolution RGB Remote Sensing Images

Cited by 2 publications

References 14 publications

Land Use Classification of High-Resolution Multispectral Satellite Images With Fine-Grained Multiscale Networks and Superpixel Postprocessing

Land Use Classification of High-Resolution Multispectral Satellite Images With Fine-Grained Multiscale Networks and Superpixel Postprocessing

Extraction of Land Information, Future Landscape Changes and Seismic Hazard Assessment: A Case Study of Tabriz, Iran

Contact Info

Product

Resources

About