Synthetic aperture radar (SAR) ship detection is a heated and challenging problem. Traditional methods are based on hand-crafted feature extraction or limited shallow-learning features representation. Recently, with the excellent ability of feature representation, deep neural networks such as faster region based convolution neural network (FRCN) have shown great performance in object detection tasks. However, several challenges limit the applications of FRCN in SAR ship detection: (1) FRCN with a fixed receptive field cannot match the scale variability of multiscale SAR ship objects, and the performance degrade when the objects are small; (2) as a two-stage detector, FRCN performs an intensive computation and leads to low-speed detection; (3) when the background is complex, the imbalance of easy and hard examples will lead to a high false detection. To tackle the above issues, we design a multilayer fusion light-head detector (MFLHD) for SAR ship detection. Instead of using a single feature map, shallow high-resolution and deep semantic feature are combined to produce region proposal. In detection subnetwork, we propose a light-head detector with large-kernel separable convolution and position sensitive pooling to improve the detection speed. In addition, we adapt focal loss to loss function and training more hard examples to reduce the false alarm. Extensive experiments on SAR ship detection dataset (SSDD) show that the proposed method achieves superior performance in SAR ship detection both in accuracy and speed.

Various type of deep convolution neural network (CNN) have been successfully applied to synthetic aperture radar (SAR) image despeckling. However, the model is limited by the length of network structure and expensive computation. In this paper, we proposed a novel deep learning-based approach learning the mapping from noisy image to noisy-free image with a dilated densely connected network (SAR-DDCN). Compared with the recently proposed despeckling methods, our method can achieve superior performance with less computational cost. Specifically, we replace standard convolution with dilated convolution, which could enlarge the receptive field and keep the size of feature maps at the same time. In addition, densely connected network are added to the network to alleviate the problem of gradient vanishing. Extensive experiments on synthetic and real SAR image show that the proposed method achieve better performance with respect to state-of-the-art-methods on both quantitative measurements and visual results.