From 77482a7a4bb143db95ff70aaea1b445ccb96db84 Mon Sep 17 00:00:00 2001 From: fd-col Date: Wed, 28 Feb 2024 12:29:29 +0100 Subject: [PATCH] Update 2023-02-26-Framework-Oil-and-Gas-industrial-sites.md --- ...-02-26-Framework-Oil-and-Gas-industrial-sites.md | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/_posts/2023-02-26-Framework-Oil-and-Gas-industrial-sites.md b/_posts/2023-02-26-Framework-Oil-and-Gas-industrial-sites.md index baa085f..f3c6ece 100644 --- a/_posts/2023-02-26-Framework-Oil-and-Gas-industrial-sites.md +++ b/_posts/2023-02-26-Framework-Oil-and-Gas-industrial-sites.md @@ -6,14 +6,15 @@ tags:

The framework created is responsible for automating the tasks of monitoring and surveillance of safety conditions within an Oil & Gas construction site. Drones are the tool used for collecting photos by capturing the current state of work at a certain moment over time. - - ![drone](/assets/img/senseFly.jpeg) +

-Through the use of [QGIS software](https://qgis.org/en/site/) and specific python classes, these images are -then processed in the pre-processing phase. It aims to create DSM profiles, that is a digital model of the surface relative -to the flight considered. The dataset for training models of one-dimensional convolutional neural networks (1D CNN) -consists of some of the flights already carried out by drones, and which present greater heterogeneity between classes. +![drone](/assets/img/senseFly.jpeg) +

+ Through the use of QGIS software and specific python classes, these images are + then processed in the pre-processing phase. It aims to create DSM profiles, that is a digital model of the surface relative + to the flight considered. The dataset for training models of one-dimensional convolutional neural networks (1D CNN) + consists of some of the flights already carried out by drones, and which present greater heterogeneity between classes.

The ultimate goal is to create models that identify three different classes along the entire extension of the track works: open-trench pipeline, closed-trench pipeline and pipe laying. Therefore, thanks to data augmentation techniques, customization