FE_TTPLA

Citation Author(s):
Deyu
An
Submitted by:
Deyu An
Last updated:
Wed, 11/01/2023 - 07:52
DOI:
10.21227/jkqx-yc88
Data Format:
Research Article Link:
License:
73 Views
Categories:
Keywords:
0
0 ratings - Please login to submit your rating.

Abstract 

The dataset is obtained through the transformation of mathematical tools and image processing techniques based on TTPLA. The original TTPLA dataset consisted of aerial wire data captured through pinhole cameras. After our conversion, we obtained the corresponding fisheye aerial wire data. It includes both the original images and annotated images, significantly reducing the annotation workload for fisheye wire data. We now make it publicly available for researchers to study and learn from.

The ones with the '35k' suffix are the simplified versions, while those without the '35k' are the full versions. Below is a comparison between the simplified version and TTPLA.

 Samples(training)Samples(validation)PL PixelsTTPLA1124107223.45MFE_TTPLA_35K2874317596.97G

 

In addition, the script for pinhole to fisheye conversion is also included in the zip file.

Instructions: 

The dataset contains fisheye images and annotated images, following the data format compatible with the MMsegmentation semantic segmentation framework. It can be directly employed within the MMsegmentation framework by defining an 'FE_TTPLA.py' file.

Dataset Files

    Files have not been uploaded for this dataset