S-CNN-Based Ship Detection From High-Resolution Remote Sensing Images

Zhang, Ruiqian; Yao, Jianhua; Zhang, Kao; Chen, Feng; Zhang, Jiadong

doi:10.5194/isprsarchives-xli-b7-423-2016

Cited by 51 publications

(9 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With these structures, high performance and high speed could be achieved in real-time applications, such as object detection from a video stream. Additionally, object proposal approaches have become more widely used in remote sensing applications, with improvements in speed and performance [29][30][31][32][33]. Detection by producing an object proposal achieved successful results, but there is a trade-off between the detection performance and processing speed according to the number of proposals produced.…”

Section: Introductionmentioning

confidence: 99%

Comparative Research on Deep Learning Approaches for Airplane Detection from Very High-Resolution Satellite Images

2020

View full text Add to dashboard Cite

Object detection from satellite images has been a challenging problem for many years. With the development of effective deep learning algorithms and advancement in hardware systems, higher accuracies have been achieved in the detection of various objects from very high-resolution (VHR) satellite images. This article provides a comparative evaluation of the state-of-the-art convolutional neural network (CNN)-based object detection models, which are Faster R-CNN, Single Shot Multi-box Detector (SSD), and You Look Only Once-v3 (YOLO-v3), to cope with the limited number of labeled data and to automatically detect airplanes in VHR satellite images. Data augmentation with rotation, rescaling, and cropping was applied on the test images to artificially increase the number of training data from satellite images. Moreover, a non-maximum suppression algorithm (NMS) was introduced at the end of the SSD and YOLO-v3 flows to get rid of the multiple detection occurrences near each detected object in the overlapping areas. The trained networks were applied to five independent VHR test images that cover airports and their surroundings to evaluate their performance objectively. Accuracy assessment results of the test regions proved that Faster R-CNN architecture provided the highest accuracy according to the F1 scores, average precision (AP) metrics, and visual inspection of the results. The YOLO-v3 ranked as second, with a slightly lower performance but providing a balanced trade-off between accuracy and speed. The SSD provided the lowest detection performance, but it was better in object localization. The results were also evaluated in terms of the object size and detection accuracy manner, which proved that large-and medium-sized airplanes were detected with higher accuracy.

show abstract

Section: Introductionmentioning

confidence: 99%

Comparative Research on Deep Learning Approaches for Airplane Detection from Very High-Resolution Satellite Images

2020

View full text Add to dashboard Cite

show abstract

“…This involves approaches based on the threshold, salient, shape and texture, statistics, transfer domain, computer vision, and deep learning methods [46]. Many studies detected inshore vessels using very high resolution (<1 m spatial resolution) satellite data for precise extraction of the objects [39,[47][48][49][50]. Currently, the majority (over 98%) of registered boats at the three harbors are motor fishing vessels (MFV) and fiber-reinforced plastic (FRP) [29].…”

Section: Methodsmentioning

confidence: 99%

Impact of COVID-19 Lockdown on the Fisheries Sector: A Case Study from Three Harbors in Western India

et al. 2021

View full text Add to dashboard Cite

The COVID-19 related lockdowns have brought the planet to a standstill. It has severely shrunk the global economy in the year 2020, including India. The blue economy and especially the small-scale fisheries sector in India have dwindled due to disruptions in the fish catch, market, and supply chain. This research presents the applicability of satellite data to monitor the impact of COVID-19 related lockdown on the Indian fisheries sector. Three harbors namely Mangrol, Veraval, and Vankbara situated on the north-western coast of India were selected in this study based on characteristics like harbor’s age, administrative control, and availability of cloud-free satellite images. To analyze the impact of COVID in the fisheries sector, we utilized high-resolution PlanetScope data for monitoring and comparison of “area under fishing boats” during the pre-lockdown, lockdown, and post-lockdown phases. A support vector machine (SVM) classification algorithm was used to identify the area under the boats. The classification results were complemented with socio-economic data and ground-level information for understanding the impact of the pandemic on the three sites. During the peak of the lockdown, it was found that the “area under fishing boats” near the docks and those parked on the land area increased by 483%, 189%, and 826% at Mangrol, Veraval, and Vanakbara harbor, respectively. After phase-I of lockdown, the number of parked vessels decreased, yet those already moved out to the land area were not returned until the south-west monsoon was over. A quarter of the annual production is estimated to be lost at the three harbors due to lockdown. Our last observation (September 2020) result shows that regular fishing activity has already been re-established in all three locations. PlanetScope data with daily revisit time has a higher potential to be used in the future and can help policymakers in making informed decisions vis-à-vis the fishing industry during an emergency situation like COVID-19.

show abstract

“…The early CNN-based object-detection methods first uses some traditional proposal generation algorithms [19][20][21][22][23] to extract region proposals, and then train a CNN model for classification [24][25][26] (and regression [15,27]) of each proposal. These proposal generation methods are usually time-consuming and in this way, the whole detection method cannot be trained and tested end-to-end.…”

Section: Encoder-decoder Network For Paired Semantic Segmentationmentioning

confidence: 99%

Ship Detection under Complex Backgrounds Based on Accurate Rotated Anchor Boxes from Paired Semantic Segmentation

Xiao

Zhou

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

It is still challenging to effectively detect ship objects in optical remote-sensing images with complex backgrounds. Many current CNN-based one-stage and two-stage detection methods usually first predefine a series of anchors with various scales, aspect ratios and angles, and then the detection results can be outputted by performing once or twice classification and bounding box regression for predefined anchors. However, most of the defined anchors have relatively low accuracy, and are useless for the following classification and regression. In addition, the preset anchors are not robust to produce good performance for other different detection datasets. To avoid the above problems, in this paper we design a paired semantic segmentation network to generate more accurate rotated anchors with smaller numbers. Specifically, the paired segmentation network predicts four parts (i.e., top-left, bottom-right, top-right, and bottom-left parts) of ships. By combining paired top-left and bottom-right parts (or top-right and bottom-left parts), we can take the minimum bounding box of these two parts as the rotated anchor. This way can be more robust to different ship datasets, and the generated anchors are more accurate and have fewer numbers. Furthermore, to effectively use fine-scale detail information and coarse-scale semantic information, we use the magnified convolutional features to classify and regress the generated rotated anchors. Meanwhile, the horizontal minimum bounding box of the rotated anchor is also used to combine more context information. We compare the proposed algorithm with state-of-the-art object-detection methods for natural images and ship-detection methods, and demonstrate the superiority of our method.

show abstract

S-CNN-Based Ship Detection From High-Resolution Remote Sensing Images

Cited by 51 publications

References 6 publications

Comparative Research on Deep Learning Approaches for Airplane Detection from Very High-Resolution Satellite Images

Comparative Research on Deep Learning Approaches for Airplane Detection from Very High-Resolution Satellite Images

Impact of COVID-19 Lockdown on the Fisheries Sector: A Case Study from Three Harbors in Western India

Ship Detection under Complex Backgrounds Based on Accurate Rotated Anchor Boxes from Paired Semantic Segmentation

Contact Info

Product

Resources

About