Measuring vehicles density and distribution in urban environments is an important task. The results of such estimation are highly required for different applications such as road lights configuration, congestion control, and also inter-vehicle data routing. This task, which is known as crowd sensing, is mostly based on smartphone-assisted sensing. However, in urban environments the multiple kinds of obstacles make it hard and mostly inaccurate especially for RoadSide Units (RSUs) low density cases. Furthermore, the assumption that all vehicles are collaborative and honest can lead to unexpected and unwanted situations. To address the above mentioned problems, we propose in this paper a trust-aware crowd sensing technique based on Unmanned Aerial Vehicle (UAV) for vehicular urban environments. Considering the real traffic information and the distribution of dishonest nodes in the network gathered by UAVs, our proposed solution provides a global view to both vehicles and RSUs, which can be used for different applications such as: finding the shortest and most trusted possible path to messages' final destinations, and also for the intelligent congestion control. Our simulation results show that our solution offers instant crowd and trust information over which in addition to the high detection ratios, also high packet delivery ratios with low network overhead are achieved.