Ship classification in terrestrial hyperspectral data

Keskin, Göksu; Schilling, Hendrik; Lenz, Andreas; Gross, Wolfgang L.; Middelmann, Wolfgang

doi:10.1117/12.2240875

Cited by 5 publications

(3 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Business analytics, decisionmaking, logistics have been searched via an advanced search engine. Machine-learning methods are being widely used in maritime transportation such as for trajectory prediction (Tu E. et al, 2018), pattern recognition (Kim and Jeong, 2015), prediction of ship pitching (Sun and Shen, 2009), fuel consumption (Gkerekos et al, 2019), ship classification (Keskin et al, 2016), maritime traffic risk assessment (Rong et al, 2019) and clustering problems. These studies can help understand the maritime transportation risks and provide information on the ships' trajectories on the main traffic roads (Rong et al, 2020).…”

Section: Machine-learning Methods In Maritime Transportationmentioning

confidence: 99%

Development of a machine-learning-based decision support mechanism for predicting chemical tanker cleaning activity

Cankaya

Tokgoz

Dağ

et al. 2021

JM2

View full text Add to dashboard Cite

Purpose This paper aims to propose a machine learning-based automatic labeling methodology for chemical tanker activities that can be applied to any port with any number of active tankers and the identification of important predictors. The methodology can be applied to any type of activity tracking that is based on automatically generated geospatial data. Design/methodology/approach The proposed methodology uses three machine learning algorithms (artificial neural networks, support vector machines (SVMs) and random forest) along with information fusion (IF)-based sensitivity analysis to classify chemical tanker activities. The data set is split into training and test data based on vessels, with two vessels in the training data and one in the test data set. Important predictors were identified using a receiver operating characteristic comparative approach, and overall variable importance was calculated using IF from the top models. Findings Results show that an SVM model has the best balance between sensitivity and specificity, at 93.5% and 91.4%, respectively. Speed, acceleration and change in the course on the ground for the vessels are identified as the most important predictors for classifying vessel activity. Research limitations/implications The study evaluates the vessel movements waiting between different terminals in the same port, but not their movements between different ports for their tank-cleaning activities. Practical implications The findings in this study can be used by port authorities, shipping companies, vessel operators and other stakeholders for decision support, performance tracking, as well as for automated alerts. Originality/value This analysis makes original contributions to the existing literature by defining and demonstrating a methodology that can automatically label vehicle activity based on location data and identify certain characteristics of the activity by finding important location-based predictors that effectively classify the activity status.

show abstract

Section: Machine-learning Methods In Maritime Transportationmentioning

confidence: 99%

Development of a machine-learning-based decision support mechanism for predicting chemical tanker cleaning activity

Cankaya

Tokgoz

Dağ

et al. 2021

JM2

View full text Add to dashboard Cite

show abstract

“…This study has showed that not only linear and fuzzy models but also machine learning algorithms are employed in NCR assessment studies. These machine learning algorithms have also been employed in different maritime domain such as ship domain analysis (Zhu et al, 2001), patern recognition of ship navigational data (J.-S. Kim and Jeong, 2015), prediction of ship pitching (Sun and Shen, 2009) and ship classification (Keskin et al, 2016). Due to the complex structure of our problem, we choose the best machine learning algorithms from the performance comparison results.…”

Section: Related Literaturementioning

confidence: 99%

Evaluating navigational risk of port approach manoeuvrings with expert assessments and machine learning

2019

View full text Add to dashboard Cite

“…Passive sensors with the highest number of imaging bands are the so-called hyperspectral (HS) sensors or spectrometers. Their airborne installation, image acquisition and analysis is already state of the art [8,11,12,17] and the data offers ground resolution in a centimeter scale. Since hyperspectral sensors are becoming smaller, their usage on drones is just upcoming [1,4].…”

Section: Introductionmentioning

confidence: 99%

Reconnaissance of coastal areas using simulated EnMAP data in an ERDAS IMAGINE environment

Becker

Schreiner

Auer

et al. 2018

Earth Resources and Environmental Remote Sensing/Gis Applications IX

Self Cite

View full text Add to dashboard Cite

Operations in a military naval context require a detailed planning and information gathering. For this purpose, remote sensing is a useful technique without in-situ survey. A collaboration of Fraunhofer IOSB (Institute of Optronics, System Technologies and Image Exploitation), DLR (German Aerospace Center), and Bundeswehr Geoinformation Centre ZGeoBw, engineered a desktop Geoinformation (GIS) plug-in to generate bathymetric charts and land-use-classes (vegetation, soil types and minerals) using hyperspectral data of coastal areas. These data are basis for further analyses like trafficability, barrier detection and change detection. To evaluate the potential of satellites launched in the near future with hyperspectral sensors onboard (EnMAP), aerial hyperspectral data (HySpex, AISA) were collected over a test site near the Wismar Bay in Germany and are used to simulate the satellite hyperspectral data with corresponding recording terms. Additionally, a field campaign was conducted at the Wismar Bay to acquire a ground truth dataset for model validation, including soil spectra and water depths as well. For generating the bathymetric charts, the WASI (Water Color Simulator) approach was adapted, which offers additional information besides water depth (e.g. dissolved matter, brightness of sand, relative amount of sea grass and other properties). Resulting bathymetric charts with a depth up to eight meters and unsupervised classifications of land cover are free of artefacts and accurate. A validation process is in progress. The engineered desktop GIS plug-in for HEXAGON ERDAS IMAGINE software was developed using the native SDK in addition with interoperable scripts like Python. The existing plug-in framework is variable and adaptable to different kind of GIS.

show abstract

Ship classification in terrestrial hyperspectral data

Cited by 5 publications

References 11 publications

Development of a machine-learning-based decision support mechanism for predicting chemical tanker cleaning activity

Development of a machine-learning-based decision support mechanism for predicting chemical tanker cleaning activity

Evaluating navigational risk of port approach manoeuvrings with expert assessments and machine learning

Reconnaissance of coastal areas using simulated EnMAP data in an ERDAS IMAGINE environment

Contact Info

Product

Resources

About