Automatic text placement on maps using deep learning keypoint detection models
(2022) In Student thesis series INES NGEM01 20221Dept of Physical Geography and Ecosystem Science
- Abstract
- Labeling the map is one of the most essential parts of the cartographic process that requires a huge time and energy. It is proven that the automation of map labeling is an NP-hard problem. There have been many research studies that tried to solve it such as rule-based methods, metaheuristics, and integer programming. However, the results achieved so far are not satisfactory and require much manual processing. In fact, many cartographic rules are hard to quantify or formulate as objective function or to include as a constraint. The purpose of this master thesis was to find a new way for text placement and introduce a method based on keypoint detection using deep learning. For this goal, a workflow is designed and consists of rasterization... (More)
- Labeling the map is one of the most essential parts of the cartographic process that requires a huge time and energy. It is proven that the automation of map labeling is an NP-hard problem. There have been many research studies that tried to solve it such as rule-based methods, metaheuristics, and integer programming. However, the results achieved so far are not satisfactory and require much manual processing. In fact, many cartographic rules are hard to quantify or formulate as objective function or to include as a constraint. The purpose of this master thesis was to find a new way for text placement and introduce a method based on keypoint detection using deep learning. For this goal, a workflow is designed and consists of rasterization of the manually labeled data, followed by data augmentation and shaping. Then, based on the experiments, the architecture and the parameters of the Stacked Hourglass Networks are determined based on the evaluated performance. The best-found architectures achieved an accuracy of 60.7%. Furthermore, with an attention mechanism, the model can achieve an accuracy of 63.1%. (Less)
- Popular Abstract
- The most information-dense pictorial artifacts created by mankind, maps are used to convey geographical geographic information to readers. Science and art have always been linked in mapping history. Cartographers use a set of design standards that use human creativity and expertise to produce distinctive map aesthetics that contain geospatial information.
designing the maps and finding the best place for the labels always needs lots of time and effort which cartographers try to decrease by automating it. In this thesis, some deep learning models have used to help map labeling automation with the best accuracy and less fault. AI helps to create the best result for designing maps.
The data that has been used are maps of London which are... (More) - The most information-dense pictorial artifacts created by mankind, maps are used to convey geographical geographic information to readers. Science and art have always been linked in mapping history. Cartographers use a set of design standards that use human creativity and expertise to produce distinctive map aesthetics that contain geospatial information.
designing the maps and finding the best place for the labels always needs lots of time and effort which cartographers try to decrease by automating it. In this thesis, some deep learning models have used to help map labeling automation with the best accuracy and less fault. AI helps to create the best result for designing maps.
The data that has been used are maps of London which are divided into more than 4000 images for each image the labels have been chosen as a key point and the algorithms try to train and validate the models. one of the most important achievements of this proposed technique is to reduce the time in comparison to the previous methods. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9103927
- author
- Khosravi Khorashad, Azin LU
- supervisor
- organization
- course
- NGEM01 20221
- year
- 2022
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- machine learning, Stacked hourglass networks, attention mechanism, CNN, geomatics
- publication/series
- Student thesis series INES
- report number
- 579
- language
- English
- id
- 9103927
- date added to LUP
- 2022-12-14 12:23:39
- date last changed
- 2022-12-14 12:23:39
@misc{9103927,
abstract = {{Labeling the map is one of the most essential parts of the cartographic process that requires a huge time and energy. It is proven that the automation of map labeling is an NP-hard problem. There have been many research studies that tried to solve it such as rule-based methods, metaheuristics, and integer programming. However, the results achieved so far are not satisfactory and require much manual processing. In fact, many cartographic rules are hard to quantify or formulate as objective function or to include as a constraint. The purpose of this master thesis was to find a new way for text placement and introduce a method based on keypoint detection using deep learning. For this goal, a workflow is designed and consists of rasterization of the manually labeled data, followed by data augmentation and shaping. Then, based on the experiments, the architecture and the parameters of the Stacked Hourglass Networks are determined based on the evaluated performance. The best-found architectures achieved an accuracy of 60.7%. Furthermore, with an attention mechanism, the model can achieve an accuracy of 63.1%.}},
author = {{Khosravi Khorashad, Azin}},
language = {{eng}},
note = {{Student Paper}},
series = {{Student thesis series INES}},
title = {{Automatic text placement on maps using deep learning keypoint detection models}},
year = {{2022}},
}