3D city model validation using GNSS signals
Grabowska, Celina (2020)
2020
Degree Programme in Information Technology, MSc (Tech)
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2020-11-05
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202008066396
https://urn.fi/URN:NBN:fi:tuni-202008066396
Tiivistelmä
3D city models provide three-dimensional representation of city areas and are used in visualization, navigation, urban planning and many others. For example, they are employed in positioning algorithms such as shadow matching that improves accuracy in urban areas. Correctness of the 3D models is essential so that validation methods are needed. GNSS observations carry information about receiver surroundings and could be used for that purpose. Furthermore, GNSS data can be crowd sourced which would provide huge amount of data with broad coverage.
GNSS signals are affected by different types of obstacles before reaching receivers. In urban areas the most common obstacles are high buildings. These obstacles can cause significant signal deterioration. They can cause signal attenuation, blockage or multipath. All those negative effects are reflected in GNSS measurements. For example, significant changes in the signal-to-noise ratio can indicate that a signal is being blocked.
On the other hand, given a 3D model of the buildings in an area, it is possible to predict the behavior of GNSS signals. It is, for example, possible to estimate whether a signal at a certain location is going to be shadowed by a building or not.
By comparing which signals are shadowed according to GNSS observations and 3D modelbased estimations, it is possible to detect errors in 3D maps. For instance, if GNSS measurements show blocked signals and corresponding 3D model-based predictions do not it can indicate lack of a building in a model. Extra elements in the model such as new buildings or added floors can be detected as well.
The goal of the thesis is to use GNSS observations for the validation of a 3D city map. Detected errors can provide information which areas need to be investigated and possibly re-mapped. Moreover, everyday use devices such as smartphones, watches, cars, etc. are equipped in GNSS receivers. It opens an opportunity for crowd sourcing. However, those low-cost, consumer-grade receivers have significantly lower quality than professional devices. Therefore, the second objective is to examine the feasibility of using smart devices as a source of GNSS data.
The thesis includes the analysis of the GNSS observations collected with smartphones and a professional receiver. The data was collected from locations with newly built buildings that were not yet included in the model. Moreover, the algorithm comparing GNSS and 3D model based predictions of blocked signals is presented. The results confirmed the validity of the concept. However, the errors in the 3D map were detected only using the professional receiver. The data collected with smartphones did not provide enough information and so was not enough to detect the errors.
GNSS signals are affected by different types of obstacles before reaching receivers. In urban areas the most common obstacles are high buildings. These obstacles can cause significant signal deterioration. They can cause signal attenuation, blockage or multipath. All those negative effects are reflected in GNSS measurements. For example, significant changes in the signal-to-noise ratio can indicate that a signal is being blocked.
On the other hand, given a 3D model of the buildings in an area, it is possible to predict the behavior of GNSS signals. It is, for example, possible to estimate whether a signal at a certain location is going to be shadowed by a building or not.
By comparing which signals are shadowed according to GNSS observations and 3D modelbased estimations, it is possible to detect errors in 3D maps. For instance, if GNSS measurements show blocked signals and corresponding 3D model-based predictions do not it can indicate lack of a building in a model. Extra elements in the model such as new buildings or added floors can be detected as well.
The goal of the thesis is to use GNSS observations for the validation of a 3D city map. Detected errors can provide information which areas need to be investigated and possibly re-mapped. Moreover, everyday use devices such as smartphones, watches, cars, etc. are equipped in GNSS receivers. It opens an opportunity for crowd sourcing. However, those low-cost, consumer-grade receivers have significantly lower quality than professional devices. Therefore, the second objective is to examine the feasibility of using smart devices as a source of GNSS data.
The thesis includes the analysis of the GNSS observations collected with smartphones and a professional receiver. The data was collected from locations with newly built buildings that were not yet included in the model. Moreover, the algorithm comparing GNSS and 3D model based predictions of blocked signals is presented. The results confirmed the validity of the concept. However, the errors in the 3D map were detected only using the professional receiver. The data collected with smartphones did not provide enough information and so was not enough to detect the errors.