The frequency and spatial extent of cyanobacterial bloom outbreaks have increased in recent years due to climate warming human activities, causing significant harm inland water ecosystems. Phycocyanin (PC), a characteristic pigment cyanobacteria, plays crucial role detecting blooms providing early warnings. However, accurately estimating PC turbid waters using remote sensing is challenging the optical complexity most relatively weak signal PC. To address this issue, an enhanced three-band...

This study proposes a model for leafy vegetable disease detection and segmentation based on few-shot learning framework prototype attention mechanism, with the aim of addressing challenges complex backgrounds problems. Experimental results show that proposed method performs excellently in both object semantic tasks. In task, achieves precision 0.93, recall 0.90, accuracy 0.91, mAP@50 mAP@75 0.90. is 0.95, 0.92, 0.92. These significantly outperforms traditional methods, such as YOLOv10...

Urban heat island (UHI) effect significantly influences the urban sustainability and health of cities varies seasonally. However, spring autumn have received less attention. Furthermore, research on long-term seasonal UHI changes impacts is insufficient. This study examines spatiotemporal dynamics gradient characteristics in spring, summer, autumn, winter Changsha, a typical subtropical “furnace city” from 2006 to 2022. (1) Spatiotemporal dynamics: The high-temperature (relatively zone zone)...

Due to the significant nonlinear intensity changes of multispectral images, automatic image feature point matching is a challenging task. This letter addresses problem and proposes novel descriptor combining structure texture information solve variations images. We first propose directional maps, i.e., response maps (DMs) binary (DBMs), which can capture common properties respectively. then use spatial pooling pattern histogram oriented gradients separately describe local region each...

Local region description of multi-sensor images remains a challenging task in remote sensing image analysis and applications due to the non-linear radiation variations between images. This paper presents novel descriptor based on combination magnitude phase congruency information local regions capture common features with changes. We first propose oriented maps (PCMs) binary (MBMs) using multi-oriented log-Gabor filters. The two feature vectors are then quickly constructed convolved PCMs...

Building change detection plays an imperative role in urban construction and development. Although the deep neural network has achieved tremendous success remote sensing image building detection, it is still fraught with problem of generating broken boundaries separation dense buildings, which tends to produce saw-tooth boundaries. In this work, we propose a feature decomposition-optimization-reorganization for detection. The main contribution proposed that performs by respectively modeling...

Spatiotemporal fusion in remote sensing plays an important role Earth science applications by using information complementarity between different data to improve image performance. However, several problems still exist, such as edge contour blurring and uneven pixels the predicted real ground image, extraction of salient features convolutional neural networks (CNNs). We propose a spatiotemporal method with edge-guided feature attention based on sensing, called STF-EGFA. First, module is used...

The quantitative retrieval of the chlorophyll-a concentration is an important remote sensing method that used to monitor nutritional status water bodies. high spatial resolution Sentinel-2 MSI and its subdivision in red-edge band highlight characteristics chlorophyll-a, which detection tool for assessing quality parameters plateau lakes. In this study, Nine Plateau Lakes Yunnan-Kweichow China were selected as study area. Using transit images situ measured data source, concentrations lakes...

The matching problem for heterologous remote sensing images can be simplified to the pseudo homologous via image translation improve performance. Among such applications, of synthetic aperture radar (SAR) and optical is current focus research. However, existing methods SAR-to-optical have two main drawbacks. First, single generators usually sacrifice either structure or texture features balance model performance complexity, which often results in textural structural distortion; second, due...

Forest canopy fuel moisture content (FMC) is a critical factor in assessing the vulnerability of specific area to forest fires. The conventional FMC estimation method, which relies on look-up tables and loss functions, cannot elucidate relationship between simulated data from tables. This study proposes novel approach for estimating by combining enhanced vegetation index (EVI) normalized difference (NDMI). method employs PROSAIL + PROGeoSAIL two-layer coupled radiation transfer model...

To address the challenge of image matching posed by significant modal differences in remote sensing images influenced snow cover, this paper proposes an innovative transformation-based method. Initially, Pix2Pix-GAN conversion network is employed to transform with cover into without reducing feature disparity between images. This facilitates extraction more discernible features for transforming problem from snow-covered snow-free Subsequently, a multi-level utilized extract descriptors...

Deep learning has achieved remarkable performance in semantically segmenting remotely sensed images. However, the high-frequency detail loss caused by continuous convolution and pooling operations uncertainty introduced when annotating low-contrast objects with weak boundaries induce blurred object boundaries. Therefore, a dual-stream network MAE-BG, consisting of an edge detection (ED) branch smooth boundary guidance (BG), is proposed. The ED designed to enhance edges that need be...

Abstract. Compare to optical sensors, Synthetic Aperture Radar (SAR) sensors can work at all time and under weather conditions. However, SAR images are less intuitive more difficult understand. To complement advantages of a technique image translation is put forward. Firstly, the concept named as remote sensing presented, set technology thinking for multi-source also given. Image understanding, object transformation representation considered three key steps translation, some specific...

Due to the differences in radiation and geometric characteristics of optical synthetic aperture radar (SAR) images, there is still a huge challenge for accurate matching. In this paper, we propose patch-matching network (PM-Net) improve matching performance SAR images. First, multi-level keypoints detector (MKD) with fused high-level low-level features presented extract more robust from Second, use two-channel structure image patch performance. Benefiting design, proposed method can directly...

The pansharpening method based on spatial injection is prone to spectral distortion due the detail amount mismatch. For this reason, paper proposes a that combines adaptive intensity component extraction and optimisation. Firstly, new for extracting I of multispectral (MS) image proposed, which decomposes each band MS panchromatic (PAN) according structure-energy mechanism through joint bilateral filter then determines weight coefficient extract by solving minimum optimisation problem...

In improving agricultural yields and ensuring food security, precise detection of maize leaf diseases is great importance. Traditional disease methods show limited performance in complex environments, making it challenging to meet the demands for modern agriculture. This paper proposes a model based on state-space attention mechanism, aiming effectively utilize spatiotemporal characteristics achieve efficient accurate detection. The introduces mechanism combined with multi-scale feature...

Spatial injection-based pansharpening methods are prone to spatial or spectral distortions in images due insufficient extraction of details and a mismatch between the amount detail information injected required amount. To this end, paper proposes method that optimizes injection. Firstly, optimize injection is proposed, is, extract high-frequency image through iterative filtering determine optimal number iterations based on global analysis method. Then, fully combine source image, detailed...

Multitemporal snow remote sensing image matching is an important data processing step for monitoring and environmental change analysis. The extensive coverage in images weakens local feature saliency, resulting significant differences between two making it difficult to obtain consistency features. This poses a great challenge tasks. To address this issue, we propose multitemporal method that considers global contextual can extract consistent features perform tasks even cases of coverage....

This paper presents an optimized descriptor method for multispectral images. The proposed is based on LGHD (Log-Gabor Histogram Descriptor)[1]. Initially, all feature points are detected from Long wave Infrared and Visible spectrum images, descripted by LGHD, then PCA (Principal Component Analysis) used to reduce the dimension of two different descriptors, finally descriptors match points. Experimental results show that approach achieves a better matching performance than LGHD.

Abstract. This paper proposes a new method for the matching of polygon features. Firstly, main points depicting shape feature are extracted with simplifying delineation and represented in Proximate-tangent Space. Secondly, constructing analogous estimate function based on polygon's property that going along edges covers minimum total area. With help constructed function, corresponding features objective found paired. After pairing all points, interpolation angles realize quick At last, two...