License Plate Detection Utilizing Synthetic Data from Superimposition
(2019) In Master Theses in Mathematical Sciences FMAM05 20191Mathematics (Faculty of Engineering)
- Abstract
- Machine learning projects are often constrained by data that is messy, scarce, sensitive or costly to produce. These are issues which could be mitigated by synthetic data.
This thesis tries to improve Swedish license plate localization in images by synthesizing images through superimposition, a process that produces data cheaply and in abundance.
A generative process algorithmically creates images of fake license plates which are glued on top of other images without any context awareness. The resulting images thus contain license plates where they normally don't exist, floating in the sky or across a person's face.
To evaluate the usefulness of this method, a machine learning algorithm called YOLOv3 is trained on synthesized data... (More) - Machine learning projects are often constrained by data that is messy, scarce, sensitive or costly to produce. These are issues which could be mitigated by synthetic data.
This thesis tries to improve Swedish license plate localization in images by synthesizing images through superimposition, a process that produces data cheaply and in abundance.
A generative process algorithmically creates images of fake license plates which are glued on top of other images without any context awareness. The resulting images thus contain license plates where they normally don't exist, floating in the sky or across a person's face.
To evaluate the usefulness of this method, a machine learning algorithm called YOLOv3 is trained on synthesized data and tested on real images containing Swedish license plates.
Our findings show that training on synthesized data by itself is almost good enough to match the performance of training on publicly available real data containing international license plates. The best results are achieved by mixing real and synthetic data. (Less) - Popular Abstract (Swedish)
- Artificiell intelligens (AI) har på senare tid fått en otrolig uppmärksamhet. Det tröttsamma arbetet med att manuellt koda in lösningar i ett program har ersatts av att lära programmen att hitta egna lösningar; detta är en AI-teknik som kallas för maskininlärning.
Sådana program tränar upp förmågan att känna igen mönster eller anomalier i data, vilket kan användas för allt från att köra bilar till att spela datorspel.
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/8977867
- author
- Harrysson, Olof LU
- supervisor
-
- Henning Petzka LU
- Ted Kronvall LU
- organization
- course
- FMAM05 20191
- year
- 2019
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- AI, machine learning, deep learning, CycleGAN, YOLOv3
- publication/series
- Master Theses in Mathematical Sciences
- report number
- LUTFMA-3380-2019
- ISSN
- 1404-6342
- other publication id
- 2019:E18
- language
- English
- id
- 8977867
- date added to LUP
- 2019-07-15 10:03:28
- date last changed
- 2025-02-20 12:25:25
@misc{8977867,
abstract = {{Machine learning projects are often constrained by data that is messy, scarce, sensitive or costly to produce. These are issues which could be mitigated by synthetic data.
This thesis tries to improve Swedish license plate localization in images by synthesizing images through superimposition, a process that produces data cheaply and in abundance.
A generative process algorithmically creates images of fake license plates which are glued on top of other images without any context awareness. The resulting images thus contain license plates where they normally don't exist, floating in the sky or across a person's face.
To evaluate the usefulness of this method, a machine learning algorithm called YOLOv3 is trained on synthesized data and tested on real images containing Swedish license plates.
Our findings show that training on synthesized data by itself is almost good enough to match the performance of training on publicly available real data containing international license plates. The best results are achieved by mixing real and synthetic data.}},
author = {{Harrysson, Olof}},
issn = {{1404-6342}},
language = {{eng}},
note = {{Student Paper}},
series = {{Master Theses in Mathematical Sciences}},
title = {{License Plate Detection Utilizing Synthetic Data from Superimposition}},
year = {{2019}},
}