The imaging process of the synthetic aperture radar system is not affected by time and weather, and can achieve all-day and all-weather imaging of the target. It has a wide detection range and generates high-resolution images. Therefore, it is widely used in the military and civilian fields. In recent years, satellites of "GF-3" and satellites of "HJ-1" were successively launched to fill the gap in synthetic aperture radar technology in China. However, synthetic aperture radar images still have shortcomings in the target detection process. Ship targets in SAR images are sparse and of various scales. The anchor box-based detection model relies too much on manually designed candidate boxes, which cannot adapt to all ship targets, and the parameters of the candidate boxes consume a lot of computing resources. The background of the synthetic aperture radar image is complex, and the ship target can easily disappear in the complex background, which leads to the missed detection of the detection model. SAR images contain a large number of small-scale ship targets, which are easily lost after multiple convolutions. Coherent speckle noise present in SAR images causes blurring of ship edges. To address the impact of the above problems, this paper proposes a detection model for pixel-level denoising and semantic enhancement. First, the pixel-level denoising module uses the prediction mask to generate the attention map of [0, 1], multiplies the feature map and the attention map pixel by pixel to achieve denoising, and optimizes the attention map using the cross-entropy loss. The denoising module can enhance the weight of the target area of the ship, suppress the weight of the non-target area, and enhance the difference between the ship target information and the background information in the feature map. Second, the semantic enhancement module enhances the semantic information contained in the feature map, and uses asymmetric convolutional layers to extract features of different dimensions, preventing candidate boxes with high IOU scores and low classification confidence from being suppressed. The transformer encoder is introduced in the semantic enhancement module to improve the context information between the ship target and the feature map, and enhance the dependency between the ship target and the image. Finally, the de-noised feature map with rich semantic information is fed into the detection head. In the public data set SSDD test, the model detection accuracy reaches 96.73%, the detection accuracy for small-scale ships reaches 96.85%, the detection accuracy for large-scale ships reaches 96.41%, the detection accuracy for distant sea scenes reaches 98.53%, and the detection accuracy for near-shore scenes reaches 90.00%. Compared with the detection effect of the current mainstream model in the SAR-Ship-Dataset dataset, the proposed model is verified to have a better detection effect. The experimental results show that the pixel-level denoising module uses the attention map to change the weight of the ship target area can better distinguish the target area from the non-target area. In the SAR image with complex background, the ship target position information is more obvious. The position information of small-scale ships is also enhanced, which solves the problem of small-scale ships loss after multi-layer convolution operations. The semantic enhancement module improves the model's ability to recognize ship targets and reduces the suppression of high-quality candidate boxes. Therefore, the model in this paper can effectively reduce the missed detection rate and false detection rate of the model.