Evaluation of Parking Space Detection From Aerial Imagery Using Convolutional Neural Networks
(2019) In Master's Theses in Mathematical Sciences FMAM05 20182Mathematics (Faculty of Engineering)
- Abstract
- In this thesis, the viability of using Convolutional Neural Networks to detect parking spaces using aerial imagery has been evaluated. Three state of the art networks have been tested - YOLOv3, RetinaNet, and Mask R-CNN. A dataset of urban parking lots and corresponding annotations was generated from scratch using a custom built GUI to annotate automatically generated images of parking lots from Open Street Map, from varying aerial imagery providers. This dataset was used to test and evaluate the different networks, and Mask R-CNN was used for a lengthy parameter tuning process, as it seemed to perform optimally of the three networks. The resulting model did not perform the best, believed to be because of the low amount of features... (More)
- In this thesis, the viability of using Convolutional Neural Networks to detect parking spaces using aerial imagery has been evaluated. Three state of the art networks have been tested - YOLOv3, RetinaNet, and Mask R-CNN. A dataset of urban parking lots and corresponding annotations was generated from scratch using a custom built GUI to annotate automatically generated images of parking lots from Open Street Map, from varying aerial imagery providers. This dataset was used to test and evaluate the different networks, and Mask R-CNN was used for a lengthy parameter tuning process, as it seemed to perform optimally of the three networks. The resulting model did not perform the best, believed to be because of the low amount of features represented in parking spaces. While results indicated that a somewhat complete solution is possible, it might not be feasible using a pure single CNN approach. (Less)
- Popular Abstract (Swedish)
- Vi har försökt hjälpa vidareutvecklingen av självkörande bilar med hjälp av att identifiera och klassificera parkeringsplatser i flygbilder. Problemet har visat sig vara svårt, och kanske inte kan lösas helt med hjälp av endast maskininlärningstekniker. Det finns dock tecken på att lösningen går mot rätt riktning, då vi kan se att systemet hittar de flesta av parkeringsplatserna som finns i exemplen, och dessutom sällan gissar fel.
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/8971684
- author
- Lindblom, Max LU and Nilsson, Carl LU
- supervisor
- organization
- alternative title
- Utvärdering av Parkeringsplatsdetektering från Flygbilder Genom Neurala Faltningsnätverk
- course
- FMAM05 20182
- year
- 2019
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Machine Learning, Autonomous Cars, Convolutional Neural Networks, Parking, Image Analysis
- publication/series
- Master's Theses in Mathematical Sciences
- report number
- LUTFMA-3376-2019
- ISSN
- 1404-6342
- other publication id
- 2019:E6
- language
- English
- id
- 8971684
- date added to LUP
- 2019-07-15 11:38:51
- date last changed
- 2019-07-15 11:38:51
@misc{8971684, abstract = {{In this thesis, the viability of using Convolutional Neural Networks to detect parking spaces using aerial imagery has been evaluated. Three state of the art networks have been tested - YOLOv3, RetinaNet, and Mask R-CNN. A dataset of urban parking lots and corresponding annotations was generated from scratch using a custom built GUI to annotate automatically generated images of parking lots from Open Street Map, from varying aerial imagery providers. This dataset was used to test and evaluate the different networks, and Mask R-CNN was used for a lengthy parameter tuning process, as it seemed to perform optimally of the three networks. The resulting model did not perform the best, believed to be because of the low amount of features represented in parking spaces. While results indicated that a somewhat complete solution is possible, it might not be feasible using a pure single CNN approach.}}, author = {{Lindblom, Max and Nilsson, Carl}}, issn = {{1404-6342}}, language = {{eng}}, note = {{Student Paper}}, series = {{Master's Theses in Mathematical Sciences}}, title = {{Evaluation of Parking Space Detection From Aerial Imagery Using Convolutional Neural Networks}}, year = {{2019}}, }