Geospatial Data Science Group

Three-dimensional (3D) maps are used extensively in a variety of applications, from air and noise pollution modelling to location-based services such as 3D mapping-aided Global Navigation Satellite Systems (GNSS), and positioning and navigation for emergency service personnel, unmanned aerial vehicles and autonomous vehicles. However, the financial cost associated with creating and updating 3D maps using the current state-of-the-art methods such as laser scanning and aerial photogrammetry are prohibitively expensive. To overcome this, researchers have proposed using GNSS signals to create 3D maps. This paper advances that family of methods by proposing and implementing a novel technique that avoids the difficult step of directly classifying GNSS signals into line-of-sight and non-line-of-sight classes by utilising edge detection techniques adapted from computer vision. This prevents classification biases and increases the range of environments in which GNSS-based 3D mapping methods can be accurately deployed. Being based on the patterns of blockage and attenuation of GNSS signals that are freely and globally available to receive by many mobile phones, makes the proposed technique a free, scalable and accessible solution. This paper also identifies some key indicators affecting data collection scalability and efficiency of the 3D mapping solution.

Our "digified" lives have provided researchers with an unprecedented opportunity to study society at a much higher frequency and granularity. Such data can have a large sample size but can be sparse, biased, and exclusively contributed by the users of the technologies. We look at the increasing importance of missing data and under-representation and propose a new perspective that considers missing data as useful data to understand the underlying reasons for missingness and that provides a realistic view of the sample size of large but under-represented data.