Lack of consideration of adequate safety measures during the design and operation of roundabouts would adversely affect their overall performance. There is limited research on roundabout safety, especially proactive safety analysis, under heterogeneous traffic scenarios. The present study assesses the safety of roundabouts using various macroscopic and microscopic indicators. Further, we have analyzed the potential crash risks for different vehicle classes considering the safety indicators. Field data were collected at four locations for two different time slots on a roundabout in India using Unmanned Aerial Vehicle-(UAV). Aerial monitoring overcomes the limitations of traditional traffic data collection methods due to its ability to capture large areas of the roundabout. The vehicles trajectories were extracted using the Traffic-Data-Extractor-(IITB). The macroscopic safety indicators considered in this study are Speed, volume, and speed dispersion, whereas the microscopic safety indicators considered are the Time to Collision-(TTC), Modified Time to Collision-(MTTC), and Post encroachment Time-(PET). Consistency and comparative analyses were applied to examine the performance of indicators. The results revealed that microscopic indicators would give a better assessment of the safety level at the roundabout as compared to macroscopic indicators. Moreover, the results indicate that the motorized two-wheelers are the most vulnerable class of vehicle in terms of safety indicators. The insights from this study can be used to improve the safety and reducing accident exposure of vehicles at roundabouts in developing countries. The findings of this study can be used for developing warning strategies in collision avoidance systems for autonomous vehicles in heterogeneous traffic conditions.