The increasing usage of police surveillance in today's society has led to rapid development of automated surveillance technologies, such as in the setting of predictive policing. Human anomaly detection in surveillance is affected by biases shaped by prior beliefs and expectations. Similarly, automated surveillance technologies, which typically are data driven, inherently contain biases in the training data. This highlights the importance of understanding the underlying causes of decision making in police surveillance, requiring a structured approach and to develop a conceptual mapping that can be read by humans and systems. Therefore, the research questions focus on what factors affect decisions, and how they can be structured to facilitate inferences about decision-making. To answer these research questions, 6 interviews with police experts, civilians and surveillance operators were conducted.  Following the methodological steps of Grounded Theory, the interviews were analyzed qualitatively, resulting in 10 cardinal categories that explain decision-making in police surveillance. Findings indicate that this decision process can be understood in terms of Social Cognitive Theory, revealing a set of key categories that influence decision making. On a high level this regards personal, behavioral and environmental factors. This knowledge is then implemented in an ontology, bringing the experts' knowledge into a format that can be shared, reused and understood across human and system boundaries.