Object Detection Method For Remote Sensing Imagery Based On Deep Learning

With the rapid development of imaging technology in the field of remote sensing,satellite and aerial sensors have provided massive high-resolution remote sensing images.Nowadays,object detection plays a vital role in the interpretation of remote sensing images,and it is of great significance to apply high-resolution remote sensing images to the national economy and military fields.In this article,object detection method for remote sensing imagery based on deep learning is used as the research topic.Specifically,we focuses on the problem of ship detection in SAR images and instance segmentation in remote sensing image.The purpose is to improve the performance of object detection in high-resolution remote sensing images and make detection results more accurate.The specific research content is as follows:Aiming at the problem of ship detection in SAR images,we have research on accurate and robust ship detection methods in high-resolution SAR images,especially for inshore and offshore scenes.A novel ship detection method based on a high-resolution ship detection network(HR-SDNet)for high-resolution SAR imagery is proposed.The HR-SDNet adopts a novel high-resolution feature pyramid network(HRFPN)to take full advantage of the feature maps of high-resolution and low-resolution convolutions for SAR image ship detection.In this scheme,the HRFPN connects high-to-low resolution subnetworks in parallel and can maintain high resolution.Next,the Soft Non-Maximum Suppression(Soft-NMS)is used to improve the performance of the Non-Maximum Suppression(NMS),thereby improving the detection performance of the dense ships.Then,we introduce the Microsoft Common Objects in Context(COCO)evaluation metrics,which provides not only the higher quality evaluation metrics average precision(AP)for more accurate bounding box regression,but also the evaluation metrics for small,medium and large targets,so as to precisely evaluate the detection performance of our method.Finally,the experimental results on the SAR Ship Detection Data Set(SSDD)and TerraSAR-X high-resolution images show that: HRFPN improves the detection performance by 4.3% compared to FPN in inshore scenes,which also proves COCO evaluation metrics are effective for SAR image ship detection.Aiming at the problem of remote sensing image instance segmentation,a novel instance segmentation approach of high-resolution remote sensing imagery based on Cascade Mask R-CNN is proposed,which is called high-quality instance segmentation network(HQ-ISNet).In this scheme,the HQ-ISNet exploits a high-resolution feature pyramid network(HRFPN)to fully utilize multi-level feature maps and maintain high-resolution feature maps for remote sensing images instance segmentation.Next,to refine mask information flow between mask branches,the instance segmentation network version 2(ISNetV2)is proposed to promote further improvements in mask prediction accuracy.Then,we construct a new,more challenging dataset based on the synthetic aperture radar(SAR)ship detection dataset(SSDD)and the Northwestern Polytechnical University very-high-resolution 10-class geospatial object detection dataset(NWPU VHR-10)for remote sensing images instance segmentation and it can be used as a benchmark for evaluating instance segmentation algorithms in the high-resolution remote sensing images.Finally,extensive experimental analyses and comparisons are performed on the SSDD and the NWPU VHR-10 dataset,thus proving that the proposed method is more accurate and effective than the existing instance segmentation algorithms in high-resolution remote sensing images.