A method for remote sensing image classification by combining Pixel Neighbourhood Similarity and optimal feature combination

Zhang, Kaili; Chen, Yonggang; Wang, Wentao; Wu, Yu-Di; Wang, Bo; Yan, Yanting

doi:10.1080/10106049.2022.2158948

Cited by 5 publications

(8 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…CFS shows moderate stability with a relatively smaller standard deviation of the feature subset size (Wald et al, 2013). Given it is a filter-based method, CFS is computationally efficient (Zhang et al, 2023).…”

Section: Filter-based Feature Selectionmentioning

confidence: 99%

“…For the latter, as an example, a wrapped‐based method can be followed by a filter‐based method (Saboori et al, 2022). While these are generically used for all data types, they have been proven to be usable for the classification of multispectral image data in recent case studies (Georganos et al, 2018; Saboori et al, 2022; Zhang et al, 2023).…”

Section: Data Preprocessing and Feature Engineeringmentioning

confidence: 99%

“…Filter‐based techniques are the simplest, most efficient, and most generic ones where they rank and select features or their subsets based on specific criteria, for example, correlation (Saboori et al, 2022; Zhang et al, 2023). The ranking bubbles up the most significant features, but this approach does not consider the model prediction or feature interaction for optimal selection.…”

Section: Data Preprocessing and Feature Engineeringmentioning

confidence: 99%

“…One of the popularly used filter‐based feature selection methods is the correlation‐based feature selection (CFS) (Georganos et al, 2018; Hall, 1999; Zhang et al, 2023). The original method uses a heuristic search strategy and a correlation assessment function to measure the importance of feature combinations.…”

Section: Data Preprocessing and Feature Engineeringmentioning

confidence: 99%

“…The neighborhood component analysis (NCA) uses a target function based on the average leave-one-out classification performance over the training input. NCA computes the feature weight and ranks them to extract feature subsets that maximize the function.3.4.2 | Correlation-based feature selectionOne of the popularly used filter-based feature selection methods is the correlation-based feature selection (CFS)(Georganos et al, 2018;Hall, 1999;Zhang et al, 2023). The original method uses a heuristic search strategy and a…”

mentioning

confidence: 99%

See 4 more Smart Citations

Multispectral data mining: A focus on remote sensing satellite images

Lim,

Sreevalsan‐Nair,

Daya Sagar

2023

WIREs Data Min & Knowl

View full text Add to dashboard Cite

This article gives a brief overview of various aspects of data mining of multispectral image data. We focus on specifically the remote sensing satellite images acquired using multispectral imaging (MSI), given the technology used across multiple knowledge domains, such as chemistry, medical imaging, remote sensing, and so on with a sufficient amount of variation. In this article, the different data mining processes are reviewed along with state‐of‐the‐art methods and applications. To study data mining, it is important to know how the data are acquired and preprocessed. Hence, those topics are briefly covered in the article. The article concludes with applications demonstrating the knowledge discovery from data mining, modern challenges, and promising future directions for MSI data mining research.This article is categorized under: Application Areas > Science and Technology Fundamental Concepts of Data and Knowledge > Knowledge Representation Fundamental Concepts of Data and Knowledge > Big Data Mining

show abstract

Section: Filter-based Feature Selectionmentioning

confidence: 99%

Section: Data Preprocessing and Feature Engineeringmentioning

confidence: 99%

Section: Data Preprocessing and Feature Engineeringmentioning

confidence: 99%

Section: Data Preprocessing and Feature Engineeringmentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Multispectral data mining: A focus on remote sensing satellite images

Lim,

Sreevalsan‐Nair,

Daya Sagar

2023

WIREs Data Min & Knowl

View full text Add to dashboard Cite

show abstract

Directions of land degradation in the greater Istanbul metropolitan area: A view from four decades

Bayraktar,

Cegielska,

Sökmen

et al. 2024

Land Degrad Dev

View full text Add to dashboard Cite

In an era of dynamic urbanization, especially in densely populated regions such as the Greater Istanbul Metropolitan Area, land degradation (LD) is a widespread and multifaceted phenomenon. Hence, there is a need for monitoring LD to track the direction of urban sprawl and respond rationally. In this context, the purpose of this study was to investigate the magnitude and directions of LD in Istanbul from 1984 to 2022. The originality of the study lies in the development of the CAI_A index, which is a quantitative measure of the degree of anthropogenic impact. This study utilized Landsat 5 and Landsat 8 satellite imagery from 1984, 2002, and 2022, employing a maximum likelihood classification method to determine land cover and land use changes. The results show that the area of LD by anthropogenic activity has increased by an average of nearly 40 percentage points in each district of Istanbul over the past four decades, significantly impacting both natural and semi‐natural areas. Furthermore, a noteworthy observation was made regarding forests in the northern part of the study area, which demonstrated stability and resilience to LD. The results of the study are important for the urban environment because they facilitate a better understanding of the impact of human activities on mitigating LD in the context of climate change, among other things. Moreover, the results identify specific areas in need of targeted monitoring and remedial actions regarding LD.Research Highlights Istanbul's urbanization has been consistent since 1984. Urban expansion primarily contributes to agricultural land degradation. Land degradation clusters near settlements and on the southern coast. State forests are crucial for preserving natural areas and resisting degradation. Forest services are threatened by fragmented urban expansion.

show abstract

Enhancing spatio-temporal environmental analyses: A machine learning superpixel-based approach

Estefania-Salazar,

Iglesias

2024

Heliyon

View full text Add to dashboard Cite

A method for remote sensing image classification by combining Pixel Neighbourhood Similarity and optimal feature combination

Cited by 5 publications

References 58 publications

Multispectral data mining: A focus on remote sensing satellite images

Multispectral data mining: A focus on remote sensing satellite images

Directions of land degradation in the greater Istanbul metropolitan area: A view from four decades

Enhancing spatio-temporal environmental analyses: A machine learning superpixel-based approach

Contact Info

Product

Resources

About