Land-Use/Land-Cover Change Detection Based on Class-Prior Object-Oriented Conditional Random Field Framework for High Spatial Resolution Remote Sensing Imagery

Shi, Sunan; Zhong, Yanfei; Zhao, Ji; Lv, Pengyuan; Liu, Yinhe; Zhang, Liangpei

doi:10.1109/tgrs.2020.3034373

Cited by 33 publications

(12 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most OBIC [8] classification algorithms use 1D features hand-crafted from picture objects (superpixels). This letter introduces a deep OBIC framework utilizing [9,10] Convolutional Neural Networks (CNNs) to extract 2D deep superpixel features [11]. Before designing the network [12], studied superpixel mask regulations before experiments, the proposed framework for better overall accuracy, coefficient, and Fmeasure is delivered by our DiCNN-4 (Double-input CNN) [13] model.…”

Section: Introductionmentioning

confidence: 99%

An Efficient Multispectral Image Classification and Optimization Using Remote Sensing Data

et al. 2022

View full text Add to dashboard Cite

A significant amount of effort and cost is required to collect training samples for remote sensing image classifications. The study of remote sensing and how to read multispectral images is becoming more important. High-dimensional multispectral images are created by the various bands that show how materials behave. The need for more information about things and the improvement of sensor resolutions have led to the creation of multispectral data with a higher size. In recent years, it has been shown that the high dimensionality of these data makes it hard to preprocess them in multiple ways. Recent research has demonstrated that one of the most crucial methods to address this issue is by adopting a variety of learning strategies. But as the data gets more complicated, these methodologies are not adequate to support. The proposed methodology shows that the classification experiment using remote sensing images indicates the maximum likelihood classifier with different deep learning models; weight vector (WV) AdaBoost and ADAM can greatly limit overfitting, and it obtains high classification accuracy. Proposed VGG16 and Inception v3 increase classification accuracy along with optimization process produce 96.08%.

show abstract

Section: Introductionmentioning

confidence: 99%

An Efficient Multispectral Image Classification and Optimization Using Remote Sensing Data

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Several studies were conducted based on dynamic changes of land use and land cover with land cover change detection and supervised classification (Akyürek et al 2018;Seyam et al 2023) and accuracy assessment using the kappa coefficient (Tewabe and Fentahun, 2020;Hussain and Karuppannan, 2023), and prediction by satellite image of landsat-7 and landsat-8 using cellular automated and Markov chains (Abijith and Saravanan, 2022;Wang et al 2021). Deep learning (DL)-based method (Song et al 2021), supervised classification, NDVI method (Pande et al 2021) including field verification and Google Earth Professional (Kamel, 2020), NDWI method (Raut et al 2020), MNDWI method (Bhattacharjee et al 2021), transition matrix method (Bagwan and Sopan, 2021), post classification matrix (Kouhgardi et al 2022; Márquez-Romance et al 2022; Das and Angadi, 2022), classprior object-oriented conditional random field (COCRF) method (Shi et al 2020), maximum likelihood classifier (MLC) method (Kumar and Jain, 2020; Saini et al 2019;Sarif and Gupta, 2022), Siamese global learning framework (Zhu et al 2022), preprocessing and classification and accuracy assessment (Thakur et al 2020;Mondal et al 2021;Mondal et al 2022), using various satellite imagery such as Landsat, MODIS, Sentinel and SPOT. The advantage of Landsat satellite data is the free accessibility of multi-temporal time series since 1972 (Lu et al 2019).…”

Section: Introductionmentioning

confidence: 99%

Un-planned Urban Growth Monitoring from 1991 to 2021 of Aizawl City, North-east India by Multi- Temporal Changes and CA-ANN Model

Lawmchullova,

Lalrinawma,

Rinkimi

et al. 2024

Preprint

View full text Add to dashboard Cite

Monitoring urban landuse and landcover (LULC) change is a crucial element in developing cities like Aizawl to improve land use planning for future smart cities. The objective of the current study is to analyze the lulc changes of Aizawl city between 1991 and 2021 using multi-date Landsat images and a cellular automata-artificial neural network (CA-ANN) model to predict future scenarios. The present study is highly essential for examining the urban expansion in a vertical hill city and the historical influence of settlement patterns along the edges of hill ranges for proper land use planning. The automatic classification of support vector machines (SVM) in-built at Orfeo tool box (OTB) modules was employed for LULC pattern classification. The land cover change method of the semi-automatic classification plugin (SCP) was used to identify the past LULC using Landsat 4, 5, 7, and 8. The future LULC was stimulated using the machine-learning approaches modules for land use change evaluation (Molusce) plugin in QGIS 2.18. Also, we highlight the factors that influence future LULC changes and the impacts of unplanned hill cities from the results of multi-criteria evaluation (MCE) and analytical hierarchical process (AHP). The study reveals that built-up areas are continuously increasing while open forest, agricultural land, and fallow land are diminishing, even in the projected land use land cover thematic layer in 2031. The built-up area has seen the highest change, from 5.98–25.8% in 1991 to 2021; the rate of increase has been 0.636 km2/yr-1 during the last 30 years. Similarly, dense forest cover also increased from 12.14–18.72% from 1991 to 2021, while other landuse landcover patterns like open forest, fallow land, and agricultural land are declining due to urban expansion. The accuracy level of Kappa coefficients was 97.30% in 1991 and 100% in the years 2001, 2011, and 2021, respectively. This study can help town planners and decision-makers with the protection of greenery and proper town planning.

show abstract

“…T HE past decade has witnessed a widespread growth of interest in image change detection [1] due to its innumerable applications in diverse disciplines including video surveillance [2], medical diagnosis [3], and remote sensing [4], [5], in which change detection in synthetic aperture radar (SAR) has attracted increasing attention in remote sensing communities [6], [7]. However, change detection using SAR images is still a challenging task due to the existence of speckle noise [8].…”

Section: Introductionmentioning

confidence: 99%

SSN: Stockwell Scattering Network for SAR Image Change Detection

Chen

Zhao

Wang

et al. 2023

IEEE Geosci. Remote Sensing Lett.

View full text Add to dashboard Cite

Recently, synthetic aperture radar (SAR) image change detection has become an interesting yet challenging direction due to the presence of speckle noise. Although both traditional and modern learning-driven methods attempted to overcome this challenge, deep convolutional neural networks (DCNNs)-based methods are still hindered by the lack of interpretability and the requirement of large computation power. To overcome this drawback, wavelet scattering network (WSN) and Fourier scattering network (FSN) are proposed. Combining respective merits of WSN and FSN, we propose Stockwell scattering network (SSN) based on Stockwell transform which is widely applied against noisy signals and shows advantageous characteristics in speckle reduction. The proposed SSN provides noise-resilient feature representation and obtains state-of-art performance in SAR image change detection as well as high computational efficiency. Experimental results on three real SAR image datasets demonstrate the effectiveness of the proposed method.

show abstract

Land-Use/Land-Cover Change Detection Based on Class-Prior Object-Oriented Conditional Random Field Framework for High Spatial Resolution Remote Sensing Imagery

Cited by 33 publications

References 49 publications

An Efficient Multispectral Image Classification and Optimization Using Remote Sensing Data

An Efficient Multispectral Image Classification and Optimization Using Remote Sensing Data

Un-planned Urban Growth Monitoring from 1991 to 2021 of Aizawl City, North-east India by Multi- Temporal Changes and CA-ANN Model

SSN: Stockwell Scattering Network for SAR Image Change Detection

Contact Info

Product

Resources

About