Deep learning detection methods use in ship remains a challenge, owing to the small scale of objects and interference from complex sea surfaces. In addition, existing rarely verify robustness their algorithms on multisensor images. Thus, we propose new improvement "you only look once" version 3 (YOLOv3) framework for marine surveillance, based synthetic aperture radar (SAR) optical imagery. First, improved choices are obtained anchor boxes by using linear scaling k-means++ algorithm. This...

To overcome the limitation of poor processing times for long-distance off-road path planning, an improved A-Star algorithm based on terrain data is proposed in this study. The planning tasks was developed to identify a feasible between start and destination map generated using digital elevation model. This study optimised two aspects: structure, retrieval strategy. First, hybrid structure minimum heap 2D array greatly reduces time complexity algorithm. Second, search strategy designed that...

SUMMARY We examined the spatial distribution pattern and meteorological drivers of dengue fever (DF) in Guangdong Province, China. Annual incidence DF was calculated for each county between 2005 2011 geographical using Moran's I statistic excess risk maps. A time-stratified case-crossover study used to investigate short-term relationship factors Southern Oscillation Index (SOI). High-epidemic areas were restricted Pearl River Delta region Han region, significant from 2006 2009 2011. Daily...

Sea ice is one of the most prominent marine disasters in high latitudes. Remote sensing technology provides an effective means for sea detection. images contain rich spectral and spatial information. However, traditional methods only focus on information or information, do not excavate feature simultaneously remote classification. At same time, complex correlation characteristics among spectra small sample problem classification also limit improvement accuracy. For this issue, paper proposes...

Sea ice is one of the typical causes marine disasters. image classification an important component sea detection. Optical data contain rich spectral information, but they do not allow to easily distinguish between ground objects with a similar spectrum and foreign same spectrum. Synthetic aperture radar (SAR) texture usually have single source. The limitation single-source that for further improvements accuracy remote sensing classification. In this paper, we propose method based on deep...

Fire is an important ecosystem process and has played a complex role in terrestrial ecosystems the atmosphere environment. Sometimes, wildfires are highly destructive natural disasters. To reduce their impact, must be detected as soon possible. However, accurate timely monitoring of challenging task due to traditional threshold methods easily suffered false alarms caused by small forest clearings, omission error large fires obscured thick smoke. Deep learning characteristics strong ability,...

Highway crack segmentation is a critical task for highway infrastructure monitoring and maintenance. While imagery from unmanned aerial vehicles (UAVs) applied to the of segmentation, it has great prospects in terms speed range. However, difficult accurately identify road cracks UAV remote sensing images, because are very narrow small, often containing only few pixels. To improve this study proposed an improved identification technique based on U-Net architecture enhanced with convolutional...

The accurate and timely identification of the degree building damage is critical for disaster emergency response loss assessment. Although many methods have been proposed, most them divide damaged buildings into two categories—intact damaged—which insufficient to meet practical needs. To address this issue, we present a novel convolutional neural network—namely, earthquake classification net (EBDC-Net)—for assessment based on post-disaster aerial images. proposed network comprises...

Sea ice is one of the causes marine disasters. The classification sea images an important part detection. labeled samples in hyperspectral image are difficult to acquire, which minor sample problems. In addition, most current methods mainly use spectral features for shallow learning, also limits further improvement accuracy. Therefore, this paper proposes a method based on spectral-spatial-joint feature with deep learning. proposed first extracts texture information by gray-level...

Automated coastline extraction from optical satellites is fundamental to coastal mapping, and sea-land segmentation the core technology of extraction. Deep convolutional neural networks (DCNNs) have performed well in semantic recent years. However, using deep learning techniques remains a challenging task, due lack benchmark dataset difficulty deciding which model use. We present comparative framework Landsat-8 OLI imagery via techniques. Three issues are investigated: (1) constructing...

High resolution stereo satellite images (HRSSIs) have the potential to provide accurate height and volume information, playing a crucial role in assessing building collapses during various natural disasters. However, time-consuming process of 3D reconstruction, inadequate vertical accuracy digital surface model (DSM), concentrated clustering buildings pose challenges for collapse assessment focused on buildings. Therefore, we present an improved approach rapid fine-grained collapses....

Surface meltwater critically impacts the Antarctic mass balance and global sea level rise. Quantifying extent of surface in Antarctica on a large scale is challenging task. Traditional methods, such as thresholding, have many limitations. We used deep learning method, U-Net with attention blocks, to automatically extract from Sentinel-2 images. inserted mechanism blocks into assign different weights all pixels channels utilize high resolution multiple The model was map water bodies images,...

When multiple synthetic aperture radar (SAR) images are stitched together, the intensity disconnects between them can have a significant impact on mosaic's quality. Many approaches focus decreasing differences while ignoring issue of image quality improvement. This study provides an algorithm for color correction and naturalness restoration with uneven luminance in order to generate high-quality mosaics. To increase illuminance component's naturalness, is first divided into reflectance...

Qinghai Lake, the largest inland saltwater lake in China, is an important water body that maintains ecological security of northeastern Tibetan Plateau. Global navigation satellite system reflectometry technology (GNSS-R) rarely used detection, especially boundary detection under circumstances space-borne. This paper attempts to implement using CYGNSS raw intermediate frequency (IF) data for first time. Firstly, second-order frequency-locked loop assist third-order phase-locked closed-loop...

Sea surface height (SSH) retrieval based on spaceborne Global Navigation Satellite System Reflectometry (GNSS-R) usually uses the GNSS-R geometric principle and delay-Doppler map (DDM). The traditional method condenses DDM information into a single scalar measure requires error model correction. In this article, idea of using machine learning methods to retrieve SSH is proposed. Specifically, two widely-used methods, principal component analysis combined with support vector regression...

Accurate recognition of circular marks is crucial for calibration, object tracking, and three-dimensional reconstruction in videogrammetry. However, most existing studies were designed under single or relatively simple scenes. When the algorithms are applied to more complex scenarios, it will result higher false detection miss-detection rate. In this article, we present a high-precision method based on novel deep learning model, Circular-MarkNet (CMNet) solve problem. The proposed network...

In response to the deficiency of detection capability traditional remote sensing means (scatterometer, microwave radiometer, etc.) for high wind speed above 25 m/s, this paper proposes a GNSS-R technique combined with machine learning method invert at sea surface. The L1-level satellite-based data from Cyclone Global Navigation Satellite System (CYGNSS), together European Centre Medium-Range Weather Forecasts (ECMWF) and National Centers Environmental Prediction (NCEP) data, constitute...

Light and color uniformity is essential for the production of high-quality remote-sensing image mosaics. Existing correction methods mainly use flexible models to express differences between multiple images impose specific constraints (e.g., gradient or contrast constraints) preserve texture information as much possible. Due these constraints, it usually difficult correct in during processing. We propose a method that can optimize luminance, contrast, difference images. In YCbCr space, this...

Accurate surface water mapping is crucial for monitoring and protecting ecosystem environments. During the past decades, a number of extraction methods have been presented significant progress has made. However, most existing research mainly focused on reducing interference factors such as clouds shadows, building mountain etc. There are relatively few studies concentrated fine body boundaries, which equal importance mapping. Therefore, in this paper, we developed novel boundary enhancement...

In lunar exploration missions, path planning for rovers using digital elevation models (DEMs) is currently a hot topic in academic research. However, research on large-scale DEMs has rarely been discussed, owing to the low time efficiency of existing algorithms. Therefore, this article, we propose fast path-planning method distributed tile pyramid strategy and an improved A <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">*</sup> algorithm. The...

In the classification of remote-sensing sea ice images, labeled samples are difficult to acquire. To adequately utilize massive number unlabeled samples, which contain abundant information, we propose a cooperative framework based on active learning (AL) and semi-supervised (SSL) for image classification. We acquire most valuable using AL make full use information contained in SSL, then conduct label consistency verification procedure further ensure quality pseudo-labeled obtained through...

A timely and accurate damage assessment of buildings after an earthquake is critical for the safety people property. Most existing methods based on classification segmentation use two-dimensional information to determine level buildings, which cannot provide multi-view damaged building, resulting in inaccurate results. According knowledge authors, there no related research using deep-learning-based 3D reconstruction method evaluation building damage. In this paper, we first applied MVS model...

GNSS-R technology for the retrieval of sea surface wind speed (SW) has gradually matured, and many research results in terms methodology accuracy have been obtained. Multisource data fusion a major trend remote sensing recent years. However, there are few algorithms SW retrieval, most them retrieve single source. Based on principle CYGNSS forward scattering HY-2B microwave scatterometer (HSCAT-B) backscattering, this paper proposes Fusion Model (FM) based HSCAT-B. For inversion using FM, is...