{"title":"A PolSAR ship detector using full combination of polarimetric channel","authors":"Ziwei Wang, Chao Wang, Hong Zhang","doi":"10.1109/BIGSARDATA.2017.8124918","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124918","url":null,"abstract":"Polarimetric Synthetic Aperture Radar (PolSAR) is an advanced sensor which can achieve echoes from the ship in four different polarization channels. Given a proper weighting vector of each channel in a combination process, the contrast between ships and background can be enhanced which is a huge benefit. In this paper, a new ship detector is proposed based on an automatic way to evaluate the right weighting vector. In order to speed up the algorithm, a simulated annealing particle swarm optimization (SA-PSO) is used to search the optimal solution in the complex space. Finally, a Fine Quad Radarsat-2 image is used between PWF and the proposed method, which proves the effectiveness and availability of the proposed method.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134344966","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"GF-3 SAR image despeckling based on non-subsampled shearlet transform","authors":"Zengguo Sun, R. Shi","doi":"10.1109/BIGSARDATA.2017.8124927","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124927","url":null,"abstract":"More and more attentions have been paid to GF-3 due to its diversified imaging mode, high resolution and extensive application. However, due to the coherent imaging system, speckle appears in GF-3 synthetic aperture radar (SAR) images and it hinders the understanding and interpretation of images seriously. As a new multi-resolution analysis tool, Shearlet has the best sparse representation and it can effectively deal with the high-resolution SAR images that have the obvious sparse features. In this paper, non-subsampled Shearlet transform is used to deal with GF-3 SAR images. Firstly, the logarithmic operation is applied to the GF-3 SAR image, and the Shearlet-transformed coefficients are obtained for different subbands and different directions. The hard-threshold algorithm is used for the transformed high frequency coefficients, and the Shearlet reconstruction is carried out. The despeckled image is obtained by making an exponential operation. The experimental results demonstrate that compared with other despeckling methods, the non-subsampled Shearlet transform can effectively suppress the speckle in homogeneous region and it has the highest capability of edge feature and strong point target preservation.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117037893","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Monitoring model of corn lodging based on Sentinel-1 radar image","authors":"Dong-Hee Han, Hao Yang, Guijun Yang, C. Qiu","doi":"10.1109/BIGSARDATA.2017.8124928","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124928","url":null,"abstract":"Based on the analysis of the polarization index of the Sentine-1 radar image before and after the lodging, the lodging classification model at the regional scale is established. The polarimetric index of Sentinel-1 radar image was extracted and analyzed according to the correlation analysis between the lodging index before and after the lodging. The degree of lodging was divided by the difference of plant height before and after lodging, and finally the lodging classification model was obtained. The results of correlation analysis showed that the optimal sensitivity index of maize plant height before and after lodging was VH and VV+VH, respectively. The final lodging classification model of the total sample point of lodging degree of classification accuracy was: mild lodging 97%, moderate lodging 100%, serious lodging 83%. The results show that Synthetic Aperture Radar (SAR) can effectively evaluate the degree of maize lodging at the regional scale.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115910810","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A partially coherent approach to polarization decomposition","authors":"Bo Wang, B. Chapron, R. Garello","doi":"10.1109/BIGSARDATA.2017.8124916","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124916","url":null,"abstract":"As high resolution as Radarsat2 SAR, it is possible to sense the sea surface as a number of capillary waves as well as foam. Thus an analytical model is considered to be reasonable for polarimetric SAR scattering matrix, based on the modelling of radar backscattering coefficient as the co-contribution of Bragg (polarized) and Scalar (unpolarized), which was proposed by Quilfen et al. [1] in 1999. The interaction between i.i.d. scatterers within each pixel area has been considered using a random walk model [2], with the assumption that the Bragg part is the vector sum of the random walk by all the Bragg scatterings. This analytical model could be iteratively resolved with our former estimate [3]. The reanalysis aim to prove the possibility of this kind of polarization decomposition for oceanic applications.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121503869","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Ship detection in SAR images based on an improved faster R-CNN","authors":"Jianwei Li, C. Qu, Jiaqi Shao","doi":"10.1109/BIGSARDATA.2017.8124934","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124934","url":null,"abstract":"Deep learning has led to impressive performance on a variety of object detection tasks recently. But it is rarely applied in ship detection of SAR images. The paper aims to introduce the detector based on deep learning into this field. We analyze the advantages of the state-of-the-art Faster R-CNN detector in computer vision and limitations in our specific domain. Given this analysis, we proposed a new dataset and four strategies to improve the standard Faster R-CNN algorithm. The dataset contains ships in various environments, such as image resolution, ship size, sea condition, and sensor type, it can be a benchmark for researchers to evaluate their algorithms. The strategies include feature fusion, transfer learning, hard negative mining, and other implementation details. We conducted some comparison and ablation experiments on our dataset. The result shows that our proposed method obtains better accuracy and less test cost. We believe that SAR ship detection method based on deep learning must be the focus of future research.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124677208","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Building damage assessment of compact polarimetric SAR using statistical model texture parameter","authors":"Yin Liu, Linlin Li, Qihao Chen, Meng Shu, Zhengjia Zhang, Xiuguo Liu","doi":"10.1109/BIGSARDATA.2017.8124923","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124923","url":null,"abstract":"Accurate building damage assessment is essential to providing decision support for disaster relief and reconstruction. This paper presents a method for building damage assessment using the statistical model texture parameter (SMTP) of compact polarimetric (CP) synthetic aperture radar (SAR). This paper uses the SMTP of CP data to describe the homogeneity of building, and using threshold to extract the collapsed buildings. Then the extent of the damage buildings is evaluated based on the proportion of collapsed buildings in the block. The C band, RADARSAT-2 data is used to present and analyze the performance of the proposed method. Compared with other method, the assessment accuracy is improved. The F1-score (harmonic mean of detection rate and false alarm rate) of serious/moderate/slight damage are 78.37%, 59.30% and 72.48%, respectively, and the overall accuracy is 71.99%.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115907336","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Water-change detection with Chinese Gaofen-3 simulated compact polarimetric SAR images","authors":"Xuefei Zhang, Hong Zhang, Chao Wang","doi":"10.1109/BIGSARDATA.2017.8124940","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124940","url":null,"abstract":"Gaofen-3(GF-3) satellite is the first Chinese civil C-band synthetic aperture radar (SAR) satellite launched by China Academy of Space Technology (CAST). The objective of this paper is to investigate the capability of GF-3 simulated compact polarimetric SAR data of π/4 and CTLR mode for water change detection. A multi-class water change detection method based on complex Wishart distribution for CP SAR data is performed and evaluated. The experimental results indicate that CP SAR can serve efficiently for water change detection.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125231333","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Xinping Deng, Jinsong Chen, Hongzhong Li, Luyi Sun
{"title":"Statistical analysis of polarimetric SAR data using log-cumulants","authors":"Xinping Deng, Jinsong Chen, Hongzhong Li, Luyi Sun","doi":"10.1109/BIGSARDATA.2017.8124935","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124935","url":null,"abstract":"Introducing knowledge of the statistical properties could improve the performance of PolSAR data applications such as speckle filtering, land-use analysis, ground cover classification. As the successful launch of the most advanced Polarimetric SAR (PolSAR) instruments including RadarSat-2, TerraSAR-X, ALOS-2, Sentinel-1, and GF3, we are faced with a vast volume of high resolution PolSAR data. To extract statistical information from the big SAR data accurately and efficiently, some simple but effective statistics are in need. It is reported that the log-cumulants are useful for PolSAR data analysis, and we can use them to design estimators for distribution parameters with low bias and variance. In this paper, log-cumulants of different statistical models are compared in a graphic manner to determine the most suitable model for different land types.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134368061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Deep highway unit network for land cover type classification with GF-3 SAR imagery","authors":"Yujuan Guo, E. Chen, Ying Guo, Zeng-yuan Li, Chonggui Li, Kunpeng Xu","doi":"10.1109/BIGSARDATA.2017.8124926","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124926","url":null,"abstract":"The fully polarized synthetic aperture radar (SAR) is an advanced earth observation system with day and night imaging capability, which can obtain rich information of terrain and has a wide range of applications in environmental protection, urban planning and resource investigation. As the first self-developed C-band multi-polarized SAR image, the acquisition of massive data and operational operation of Chinese SAR remote sensing has entered the era of big data. Under the era of remote sensing large data, however, SAR image interpretation is a great challenge for scientific applications. At present, big data-based intelligent methods such as computer vision technology have achieved great success. Deep learning such as deep highway unit networks has revolutionized the computer vision area. However, due to the characteristics of SAR microwave band imaging and phase coherence processing, SAR images are very different from ordinary optical images in terms of band, projection direction, data composition and so on. Therefore, deep learning can not be directly used for quad-pol SAR image classification. In this paper, deep learning is applied to land cover type classification with GF-3 quad-pol SAR imagery. A deep highway unit network is employed to automatically extract a hierarchic feature representation from the data, based on which the land cover type classification can be conducted. Our classification model is trained on limited training data from forest resource inventory and planning data, and tested on a Radarsat-2 quad-pol images, which is the image of the same area acquired at different times. We also employ the machine learning such as SVM, Random Forest on the same samples for comparison. The deep highway unit network trained by the GF-3 images, which can reduce speckle, fully excavate the regularity of SAR images in time and space.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"210 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132696256","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Combining single shot multibox detector with transfer learning for ship detection using Sentinel-1 images","authors":"Yuanyuan Wang, Chao Wang, Hong Zhang","doi":"10.1109/BIGSARDATA.2017.8124924","DOIUrl":"https://doi.org/10.1109/BIGSARDATA.2017.8124924","url":null,"abstract":"Synthetic aperture radar (SAR) images are all-weather, all-time, and wide coverage, increasingly used for ship detection to ensure marine surveillance and transportation security. Currently, deep learning has achieved enormous success in object detection with the capability of representation learning. Combining single shot multiBox detector (SSD) with transfer learning is proposed to address ship detection with complex surroundings, such as both ocean and island in this paper. SSD is chosen because its detection accuracy remains high with relative fast speed and transfer learning is chosen because it performs well even with small training datasets. Two types of S SD models integrated with transfer learning, namely, SSD300 and SSD512 with an input size of 300 pixels and 512 pixels in height and width, are applied to ship detection. To evaluate our approach, SAR images dataset acquired by Sentinel-1 are used. Experimental results reveal that compared with SSD300, SSD512 achieves lower false alarm and slighter lower in detection accuracy. These results demonstrate the effectiveness of our method.","PeriodicalId":334489,"journal":{"name":"2017 SAR in Big Data Era: Models, Methods and Applications (BIGSARDATA)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129248802","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}