High-resolution image processing for land surface monitoring is fundamental to analyze the impact ofdifferent geomorphological processes on Earth surface for different climate change scenarios. In thiscontext, photogrammetry is one of the most reliable techniques to generate high-resolutiontopographic data, being key to territorial mapping and change detection analysis of landforms inhydro-geomorphological high-risk areas. An important issue arises as soon as the main goal is toconduct analyses over extended areas of the Earth surface (such as fluvial systems) in a short time,since the need to capture large datasets to develop detailed topographic models may limit thephotogrammetric process, due to the high demand of high-performance hardware. In order toinvestigate the best set up of computing resources for these very peculiar tasks, a study of theperformance of a photogrammetric workflow based on a FOSS (Free Open-Source Software) SfM(Structure from Motion) algorithm using different cluster configurations was conducted, leveragingthe computing power of ReCaS-Bari data center infrastructure, which hosts several services such asHTC, HPC, IaaS, PaaS. Exploiting the high-computing resources available at clusters and choosingspecific set up for the workflow steps, an important reduction of several hours in the processing timewas recorded, especially compared to classic photogrammetric programs processed on a singleworkstation with commercial softwares. The high quality of the image details can be used for landcover classification and preliminary change detection studies using Machine Learning techniques. Asubset of the datasets used for the workflow implementation has been considered to test theperformance of different Convolutional Neural Networks, using progressively more complex layersequences, data augmentation and callback functions for training the models. All the results are givenin terms of model accuracy and loss and performance evaluation.
HIGH RESOLUTION IMAGE PROCESSING AND LAND COVER CLASSIFICATION FOR HYDRO- GEOMORPHOLOGICAL HIGH-RISK AREA MONITORING
Miniello, G.;La Salandra, M.
2021-01-01
Abstract
High-resolution image processing for land surface monitoring is fundamental to analyze the impact ofdifferent geomorphological processes on Earth surface for different climate change scenarios. In thiscontext, photogrammetry is one of the most reliable techniques to generate high-resolutiontopographic data, being key to territorial mapping and change detection analysis of landforms inhydro-geomorphological high-risk areas. An important issue arises as soon as the main goal is toconduct analyses over extended areas of the Earth surface (such as fluvial systems) in a short time,since the need to capture large datasets to develop detailed topographic models may limit thephotogrammetric process, due to the high demand of high-performance hardware. In order toinvestigate the best set up of computing resources for these very peculiar tasks, a study of theperformance of a photogrammetric workflow based on a FOSS (Free Open-Source Software) SfM(Structure from Motion) algorithm using different cluster configurations was conducted, leveragingthe computing power of ReCaS-Bari data center infrastructure, which hosts several services such asHTC, HPC, IaaS, PaaS. Exploiting the high-computing resources available at clusters and choosingspecific set up for the workflow steps, an important reduction of several hours in the processing timewas recorded, especially compared to classic photogrammetric programs processed on a singleworkstation with commercial softwares. The high quality of the image details can be used for landcover classification and preliminary change detection studies using Machine Learning techniques. Asubset of the datasets used for the workflow implementation has been considered to test theperformance of different Convolutional Neural Networks, using progressively more complex layersequences, data augmentation and callback functions for training the models. All the results are givenin terms of model accuracy and loss and performance evaluation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
