Detecting trees near electric wires with Deep Learning

Published: 11-01-2021| Version 1 | DOI: 10.17632/zw4kstvxwp.1
Contributor:
Artur André Oliveira

Description

Urban images and scripts used to train and test a deep leaning framework architecture to classify images as having or not intersections between trees and overhead powerlines/wires.

Files

Steps to reproduce

The deep learning network scripts, weights, and other configuration files can be found at the `Deep Learning Folder`. In that folder, after installing the required packages (e.g. with `pip install -r requirements.txt') one can execute the `test_model.py` script to train and test the provided network named `mobilenetv2_test39`. The `Test_Data` and `Training_Data` folders contain the images and label files (summan.smr) to be executed with the test_model.py script.