• N.M Bhagya Sree
• S. Subhash Kumar

On both detection and ranging (SAR) and optical satellite imaging, the detection of marine and onshore river ships has been examined. Traditional ship recognition techniques usually on SAR photos, on the other hand, might have a high false alarm rate and be impacted by the sea level model, particularly in streams and offshore places. On tiny and accumulating ships, traditional detection techniques based on optical pictures do not function well. The idea of neural architectures is used in this research to provide a rapid geographic deep convolution network (R-CNN) technique for detecting ships in high-resolution satellite data. To begin, we select GaoFen-2 optical satellite photos with a horizontal resolution m and use a R-R-CNN to partition the large image region The area is divided into discrete fields of interest (ROI) that could contain ships. The ROI photographs are then subjected to ship recognition techniques rely on a geographical area deep neural network (R-CNN). We use an effective option approach to detect, Faster-RCNN, and keep improving the architecture of its previous convolutional (CNN), VGG16, while using delivery is available feature representations and having to perform ROI consolidation on a larger previous layer in a proposed method to achieve more accurate result of comparatively tiny and collecting ships (RPN). Finally, we make a comparison the element ( fe model (DPM), a further two widely used target recognition architectures, the bolt action fully convolutional analyzer , the previous VGG16-based Faster-RCNN, and our enhanced Faster-RCNN to one of the most impactful classic ship detection technique, the active shape model . Experiments show that our revised Faster-RCNN approach outperforms the competition.

Ship detection, RCNN, CNN

"SHIP DETECTION FROM SATELLITE IMAGES USING DEEP LEARNING", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 6, page no.j37-j43, June-2022, Available :http://www.jetir.org/papers/JETIR2206906.pdf

"SHIP DETECTION FROM SATELLITE IMAGES USING DEEP LEARNING", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 6, page no. ppj37-j43, June-2022, Available at : http://www.jetir.org/papers/JETIR2206906.pdf