An Adaptive Polygonal Centroidal Voronoi Tessellation Algorithm for Segmentation of Noisy Sar Images

Askari, Ghasem; Li, Y.; MoezziNasab, R.

doi:10.5194/isprsarchives-xl-2-w3-65-2014

“…A negative number indicates the classification is significantly worse than random. A value close to 1 indicates that the classification is significantly better than random [ 39 , 40 ]. In this analysis, accuracy assessment was performed by comparing the map produced by remote sensing analysis and the reference geological map of the study area.…”

Section: Methodsmentioning

confidence: 99%

Band Ratios Matrix Transformation (BRMT): A Sedimentary Lithology Mapping Approach Using ASTER Satellite Sensor

Askari

¹

,

Pour

²

,

Pradhan

³

et al. 2018

Sensors

View full text Add to dashboard Cite

Remote sensing imagery has become an operative and applicable tool for the preparation of geological maps by reducing the costs and increasing the precision. In this study, ASTER satellite remote sensing data were used to extract lithological information of Deh-Molla sedimentary succession, which is located in the southwest of Shahrood city, Semnan Province, North Iran. A robust and effective approach named Band Ratio Matrix Transformation (BRMT) was developed to characterize and discriminate the boundary of sedimentary rock formations in Deh-Molla region. The analysis was based on the forward and continuous division of the visible-near infrared (VNIR) and the shortwave infrared (SWIR) spectral bands of ASTER with subsequent application of principal component analysis (PCA) for producing new transform datasets. The approach was implemented to ASTER spectral band ratios for mapping dominated mineral assemblages in the study area. Quartz, carbonate, and Al, Fe, Mg –OH bearing-altered minerals such as kaolinite, alunite, chlorite and mica were appropriately mapped using the BRMT approach. The results match well with geology map of the study area, fieldwork data and laboratory analysis. Accuracy assessment of the mapping result represents a reasonable kappa coefficient (0.70%) and appropriate overall accuracy (74.64%), which verified the robustness of the BRMT approach. This approach has great potential and capability for mapping sedimentary succession with diverse local–geological–physical characteristics around the world.

show abstract

“…Then, a robust soil sampling method for a supervised k-means clustering classification will be employed. In essence, K-means clustering analysis is a machine learning statistical method where entities are grouped into categorized clusters given shared factors (Askari et al, 2014;Kim et al, 2021) This classification is conducted on the natural colour bands of the electromagnetic spectrum, as well as statistical indices such as Principal Component Analysis (PCA). Lastly, a conditional generative adversarial network (cGAN) will be trained and evaluated for is efficacy in generating soil classification maps from natural colour images, which will allow analysts to forgo the k-means clustering classification step, and employ the model in novel areas, saving time, and, potentially, financial capital, within the realm of agricultural analysis.…”

Section: Research Goalsmentioning

confidence: 99%

Soil Analysis via Remote Sensing and Artificial Intelligence for Precision Regenerative Agriculture

Kemp

2024

Preprint

0

View full text Add to dashboard Cite

<p>Soil electrical conductivity maps were generated for greenspace in the Greater Toronto Area using a conditional generative adversarial network (CGAN), which is a form of deep learning where one neural network is used to train another. The results of the analysis show that the model can accurately predict soil conductivity 34.6% of the time. It could possibly be strengthened with the inclusion of more electromagnetic bands in the supervised classifications used to train the network, such as the infrared spectrum, as well as Light Detection and Ranging (LiDAR) data. This three-dimensional imagery should be considered, as the model is not optimized when soil is obscured by foliage. Generally, these two data types are commercially available and commonly used for the analysis of greenspace. Microdrones can potentially be equipped with computer vision-enabled sensors operating this neural model to iteratively analyze soil types and complete aerial cropping.</p>

show abstract

“…Then, a robust soil sampling method for a supervised k-means clustering classification will be employed. In essence, K-means clustering analysis is a machine learning statistical method where entities are grouped into categorized clusters given shared factors (Askari et al, 2014;Kim et al, 2021) This classification is conducted on the natural colour bands of the electromagnetic spectrum, as well as statistical indices such as Principal Component Analysis (PCA). Lastly, a conditional generative adversarial network (cGAN) will be trained and evaluated for is efficacy in generating soil classification maps from natural colour images, which will allow analysts to forgo the k-means clustering classification step, and employ the model in novel areas, saving time, and, potentially, financial capital, within the realm of agricultural analysis.…”

Section: Research Goalsmentioning

confidence: 99%

Soil Analysis via Remote Sensing and Artificial Intelligence for Precision Regenerative Agriculture

Kemp

2024

Preprint

0

View full text Add to dashboard Cite

<p>Soil electrical conductivity maps were generated for greenspace in the Greater Toronto Area using a conditional generative adversarial network (CGAN), which is a form of deep learning where one neural network is used to train another. The results of the analysis show that the model can accurately predict soil conductivity 34.6% of the time. It could possibly be strengthened with the inclusion of more electromagnetic bands in the supervised classifications used to train the network, such as the infrared spectrum, as well as Light Detection and Ranging (LiDAR) data. This three-dimensional imagery should be considered, as the model is not optimized when soil is obscured by foliage. Generally, these two data types are commercially available and commonly used for the analysis of greenspace. Microdrones can potentially be equipped with computer vision-enabled sensors operating this neural model to iteratively analyze soil types and complete aerial cropping.</p>

show abstract

An Adaptive Polygonal Centroidal Voronoi Tessellation Algorithm for Segmentation of Noisy Sar Images

Cited by 7 publications

References 10 publications

Band Ratios Matrix Transformation (BRMT): A Sedimentary Lithology Mapping Approach Using ASTER Satellite Sensor

Band Ratios Matrix Transformation (BRMT): A Sedimentary Lithology Mapping Approach Using ASTER Satellite Sensor

Soil Analysis via Remote Sensing and Artificial Intelligence for Precision Regenerative Agriculture

Soil Analysis via Remote Sensing and Artificial Intelligence for Precision Regenerative Agriculture

Contact Info

Product

Resources

About