Streets are one of the most significant elements of urban spaces to accommodate public activity and provide access to numerous locations and services.  Today, our cities face various challenges, including urban sprawl, air pollution, and obesity. To address some of these problems, cities and local governments consider land use planning and urban design to lower car dependency and promote walking and biking. Many studies have characterized the built environment in terms of D variables density, diversity, design, destination accessibility, and distance to transit. Measurement of four D variables is relatively straightforward, while the D variable, design, is more nuanced.  Urban designers believe that micro features are essential for active street life but have little empirical evidence to back the claim. The conventional research approaches to objectively measure the subjective qualities of the street environment require resource-intensive street-level field observation believed to be one of the main constraints of the widespread application of such methods. This dissertation consisted of two studies that collectively explored the applications of advanced technologies and methods such as Google Street View, computer vision, and deep learning toward automating the measuring of micro-scale characteristics of the built environment related to pedestrian activity and overall walkability.