Object detection is a key technology for intelligent interpretation of remote sensing (RS) images. Existing object detection methods were primarily designed for nadir-imaging satellite RS images. Additionally, their insufficient utilization of imaging prior information makes them struggle to effectively address the significant object scale variations caused by UAV oblique imaging. To address the demand for high-accuracy object detection under oblique imaging perspectives, an imaging prior-driven super-resolution (IPSR) approach was proposed. An imaging distance model was established based on multiple parameters including camera focal length, flight altitude, and pitch angle, enabling the estimation of imaging distances at arbitrary positions in the image. Then, full-image adaptive SR reconstruction was performed based on distance estimation, effectively mitigating intra-image scale disparities under oblique imaging perspectives. Experiments on the UAV multi-view imaging dataset VSAI and satellite imagery dataset DOTA-v1.5 demonstrate that IPSR can be integrated with any mainstream object detection model. It significantly enhances the detection accuracy of these models under oblique imaging conditions (achieving gains of 6-7 mAP on VSAI and 3-4 mAP on DOTA-v1.5). Besides, it exhibits the versatility of the proposed method across multiple scenarios.