Power transmission line dataset

Citation Author(s):
Lucas
Diniz
Tiago
Santa Maria
Guilherme Augusto
Pussente
Submitted by:
Lucas Diniz
Last updated:
Thu, 02/17/2022 - 15:13
DOI:
10.21227/t9qk-cn48
Data Format:
License:
0
0 ratings - Please login to submit your rating.

Abstract 

This dataset is composed by both real and sythetic images of power transmission lines, which can be fed to deep neural networks training and applied to line's inspection task. The images are divided into three distinct classes, representing power lines with different geometric properties. The real world acquired images were labeled as "circuito_real" (real circuit), while the synthetic ones were identified as "circuito_simples" (simple circuit) or "circuito_duplo" (double circuit). There are 348 total images for each class, 232 inteded for training and 116 aimed for validation/testing.

Instructions: 

This dataset contains 1044 images, 348 acquired in a real-world scenario and 696 generated from a virtual, sythetic enviroment. There are three distinct classes, namely "circuito_real", "circuito_simples" and "circuito_duplo", equally represented in the dataset (348 samples each).  232 images of each class are inteded for training, while the remaining 116 can be splitted in validation/testing purposes.

The first class, "circuito_real" or real circuit, is composed by real-world acquired images, where wires are disposed in a sometimes messier arrangement. The other two classes are constituded of sythentic images, where "circuito_simples" or simple circuit samples can be distinguished by a central line between four extreme wires (two in each side) and "circuito_duplo" or double circuit samples presents four extreme lines each side, without a central line.

Each image name contains its label, purpose (training or validation/testing) and numeration according to these two characteristhics.