Scale in Scale for SAR Ship Instance Segmentation

نویسندگان

چکیده

Ship instance segmentation in synthetic aperture radar (SAR) images can provide more detailed location information and shape information, which is of great significance for port ship scheduling traffic management. However, there little research work on SAR segmentation, the general accuracy low because characteristics target task, such as multi-scale, aspect ratio, noise interference, are not considered. In order to solve these problems, we propose an idea scale (SIS) segmentation. Its essence establish multi-scale modes a single scale. consideration characteristic targeted SIS equipped with four tentative this paper, i.e., input mode, backbone RPN mode (region proposal network), ROI interest). The establishes inputs enhances ability extract features. makes bounding boxes better accord ratios. expands receptive field. Combined them, network (SISNet) reported, dedicated high-quality basis prevailing Mask R-CNN framework. For R-CNN, also redesign (1) its feature pyramid (FPN) small detection (2) head (DH) refined box regression. We conduct extensive experiments verify effectiveness SISNet open SSDD HRSID datasets. experimental results reveal that surpasses other nine competitive models. Specifically, average precision (AP) index superior suboptimal model by 4.4% 2.5% HRSID.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Automatic Ship Detection in SAR Images Using Multi-Scale Heterogeneities and an A Contrario Decision

The robust detection of ships is one of the key techniques in coastal and marine applications of synthetic aperture radar (SAR). Conventional SAR ship detectors involved multiple parameters, which need to be estimated or determined very carefully. In this paper, we propose a new ship detection approach based on multi-scale heterogeneities under the a contrario decision framework, with a few par...

متن کامل

simulation and experimental studies for prediction mineral scale formation in oil field during mixing of injection and formation water

abstract: mineral scaling in oil and gas production equipment is one of the most important problem that occurs while water injection and it has been recognized to be a major operational problem. the incompatibility between injected and formation waters may result in inorganic scale precipitation in the equipment and reservoir and then reduction of oil production rate and water injection rate. ...

Segmentation in Scale Space

A segmentation scheme based on tracing objects and borders through scale space is proposed. Scale space allows to create a hierarchical representation of input data which can be used to tessellate input space into objects with closed and orientable borders. For analyzing the structure of scale space, a neural network approach using synchronizing neural oscillators is proposed.

متن کامل

Multi{Scale Segmentation of Grey{Scale Images

The topic of this thesis is segmentation of grey{scale images by accessing and relating the structures of the image for all reasonable scales. Grey{scale images are regarded as physical measurements obtained by an ensemble of detectors with nite aperture and nite readout precision. Based on this point of view, the notion of image is established using linear Gaussian scale space and Morse theory...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

ژورنال

عنوان ژورنال: Remote Sensing

سال: 2023

ISSN: ['2315-4632', '2315-4675']

DOI: https://doi.org/10.3390/rs15030629