Synthetic aperture radar (SAR) image detection is an advanced remote sensing technology that acquires high-resolution images of the ground surface or target objects by transmitting and receiving radar waves. To date, many experiments and studies related to SAR images have been conducted, however, the detection of targets in SAR images is an important topic. Nevertheless, owing to the imaging principle of SAR images, it is difficult to avoid the introduction of noise in the imaging process, which creates a considerable obstacle to the detection of SAR images. Although many models for SAR image detection exist, most of them are focused on detection network enhancement, few studies can combine filtering and detection to solve the SAR image detection problem. In this paper, we offer a new filtering module, which consists of two filtering mechanisms and a fusion mechanism, which can filter out the low-level interference noise and high-level interference noise respectively, similarly, it can be added to different detectors to enhance the detection performance of the model. The experimental results show that our proposed MLF filtering module can significantly improve the detection performance of the target detection model. For the common dataset and detection model, compared with the detection without the filtering module, the detection model with the MLF improves the mAP value by 5.34% in the inshore test, 0.36% in the offshore test, and 1.58% in the comprehensive test.