The widespread availability of small-scale, off-the-shelf Unmanned Aerial Vehicles (UAVs) has opened various avenues for drone-based photography and videography, where limited human expertise is needed to carry out drone flying and post-processing of the acquired media. However, these off-the-shelf, closed-loop systems lack the flexibility and range of sensors/metadata to perform specialized operations, such as accurate mapping for Geographical Information Systems (GIS) applications. To achieve survey-grade precision, often larger UAV platforms are required, which can carry more sophisticated payloads leading to an increase in the overall cost of the systems while introducing sensor calibration complexities. To address these issues, we present an efficient approach using progressive, sequence-based feature matching and image alignment to generate high-quality, seamless, orthorectified mosaics using low-cost UAVs equipped with general-purpose imaging and positioning sensors. Our proposed approach does not require any prior knowledge of the environment or specific flight planning protocols. We validate the proposed approach through real-world, outdoor experiments by collecting and processing flight data and generating geo-referenced orthomosaics. The results showcase that our approach delivers operationally acceptable accuracy while achieving a mosaicking speed of 2.3 frames per second using consumer-grade UAV platforms with minimal sensor data. The proposed solution aims to enable rapid and cost-effective, UAV-based mapping for applications in forestry, hydrology, urban planning and disaster management.