The paper introduces SARDet-100K, a new benchmark dataset for large-scale Synthetic Aperture Radar (SAR) object detection, and the Multi-Stage with Filter Augmentation (MSFA) pretraining framework. The SARDet-100K dataset is created by merging 10 existing SAR detection datasets, resulting in a dataset with approximately 117K images and 246k instances across six categories, making it the first COCO-level large-scale multi-class SAR object detection dataset. The MSFA framework addresses the challenges of domain and model gaps in SAR object detection by employing handcrafted features and a multi-stage pretraining strategy. The framework enhances the performance of SAR object detection models and demonstrates excellent generalizability and flexibility across various deep network models. The paper also includes experimental results showing the effectiveness of the proposed methods and comparisons with state-of-the-art techniques. The dataset and code are available for public use, aiming to advance research in SAR object detection.The paper introduces SARDet-100K, a new benchmark dataset for large-scale Synthetic Aperture Radar (SAR) object detection, and the Multi-Stage with Filter Augmentation (MSFA) pretraining framework. The SARDet-100K dataset is created by merging 10 existing SAR detection datasets, resulting in a dataset with approximately 117K images and 246k instances across six categories, making it the first COCO-level large-scale multi-class SAR object detection dataset. The MSFA framework addresses the challenges of domain and model gaps in SAR object detection by employing handcrafted features and a multi-stage pretraining strategy. The framework enhances the performance of SAR object detection models and demonstrates excellent generalizability and flexibility across various deep network models. The paper also includes experimental results showing the effectiveness of the proposed methods and comparisons with state-of-the-art techniques. The dataset and code are available for public use, aiming to advance research in SAR object detection.