High-resolution deep-learning-based remote-sensing imagery analysis has been widely used in land-use and crop-classification mapping. However, the influence of composite feature bands, including complex indices arising from different sensors on backbone, patch size, predictions transferable deep models require further testing. The experiments were conducted six sites Henan province 2019 to 2021. This study sought enable transfer classification across regions years for Sentinel-2A (10-m...

As tea is an important economic crop in many regions, efficient and accurate methods for remotely identifying plantations are essential the implementation of sustainable practices periodic monitoring. In this study, we developed tested a method plantation identification based on multi-temporal Sentinel-2 images multi-feature Random Forest (RF) algorithm. We used phenological patterns cultivation China's Shihe District (such as multiple annual growing, harvest, pruning stages) to extracted...

Recent advances in video compression introduce implicit neural representation (INR) based methods, which effectively capture global dependencies and characteristics of entire sequences. Unlike traditional deep learning approaches, INR-based methods optimize network parameters from a perspective, resulting superior potential. However, most current INR utilize fixed uniform architecture across all frames, limiting their adaptability to dynamic variations within between This often leads...

In this study, MODIS normalized difference vegetation index (NDVI), TRMM3B43 precipitation, and MOD11A2 land-surface temperature (LST) data were used as sources in an analysis of temporal spatial characteristics changes ecological environmental quality the Huaihe River basin, China, from 2003 to 2018. The Mann–Kendall (MK) non-parametric test Theil–Sen slope combined for analysis; then, when with results MK mutation two introduced indexes, kurtosis coefficient (KU) skewness (SK) correlations...

Accurate acquisition of cultivated land area and location information is great significance to agricultural management, agro-ecological environment monitoring, national food security. The rapid development deep learning technology provides a new way extract information. However, there are many parameters involved in learning, so it time-consuming find the optimal parameters. In order simplify complex parameter tuning process explore main that affect classification accuracy this study uses...

Existing pedestrian attribute recognition (PAR) algorithms are mainly developed based on a static image. However, the performance is not reliable for images with challenging factors, such as heavy occlusion, motion blur, etc. In this work, we propose to understand human attributes using video frames that can make full use of temporal information. Specifically, formulate video-based PAR vision-language fusion problem and adopt pre-trained big models CLIP extract feature embeddings given...

Most existing clustering methods require prior knowledge, such as the number of clusters and thresholds. They are difficult to determine accurately in practice. To solve problem, this study proposes a novel algorithm named GEP-Cluster based on Gene Expression Programming (GEP) without knowledge. The main contributions include: (1) new concept Clustering Algebra is proposed that makes algebraic operation , (2) find best information automatic by GEP discover solution any (3) an AMCA (Automatic...

AIM:To investigate the spatial distribution patterns of anorectal atresia/stenosis in China. METHODS:Data were collected from Chinese Birth Defects Monitoring Network (CBDMN), a hospitalbased congenital malformations registry system.All fetuses more than 28 wk gestation and neonates up to 7 d age hospitals within monitoring sites CBDMN monitored 2001 2005.Two-dimensional graph-theoretical clustering was used divide into different clusters according average incidences sites. RESULTS:The...

The upper Huaihe River is the water-producing area of Basin and major grain oil-producing in China. changing global climate over recent years has increased frequency extreme weather reaches River. Research on responses surface water bodies to climates become increasingly important. Based all utilizable Landsat 4–8 T1–SR data mapping, spatio-temporal extraction its response were studied. We generated high-precision maps water, a comparison cartographic accuracy evaluation indices spatial...

Understanding how brain functions has been an intriguing topic for years. With the recent progress on collecting massive data and developing advanced technology, people have become interested in addressing challenge of decoding wave into meaningful mind states, with many machine learning models algorithms being revisited developed, especially ones that handle time series because nature waves. However, these models, like HMM hidden state discrete space or State Space Model continuous space,...

Layout features become highly susceptible to lithography process fluctuations due the widening subwavelength gap. Problematic layout patterns incur poor printability even if they pass design rule checking. These hotspots should be detected and removed at early phases improve manufacturability. While existing studies mainly focus on hotspot detection pattern classification, library generation is rarely addressed in literature but crucial effectiveness efficiency of detection. For an advanced...

Quantitative studies of the multiple factors influencing mountain-mass effect, which causes higher temperatures in mountainous than non-mountainous regions, remain insufficient. This study estimated air temperature Yellow River Basin, spans three different elevation ranges, using multi-source data to address uneven distribution regional meteorological stations. The differences effect for geomorphic regions at same altitude were then compared. Manner–Kendall nonparametric test was used...

According to the high-dimensional sparse features on storage of textual document, and defects existing in clustering methods or hybrid which have already been studied by now some other problems.So an improved text method based model, that is a approach (short for TGSOM-FS-FKM) tree-structured growing self-organizing maps (TGSOM) Fuzzy K-Means (FKM) proposed.The has optimized result through three times clustering.It firstly makes preprocess texts, filters majority noisy words using...

This paper proposes a algorithm for detecting manual blur on images, which is usually used to remove obvious traces when tamper images. The first blurs the test image and blocks both blurred image. Then extracts compares sharp edge points in contourlet domain of two so as detect suspicious blocks. Furthermore, differences between defocus can be indicated by our proposed method, we find out whether has been blurred. We establish rich set experimental results show that average accurate...

Digital signal processing is one of the most important major specialty foundation courses in field communication engineering and electronic information specialty. According to content course, project teaching method adopted combine spectrum analysis with software simulation. Projects unit impulse response design digital filter are combined factual application closely, which enable students learn different techniques meet practical needs. This beneficial help develop innovation spirit...

Human action recognition is an increasingly important research topic in the fields of video sensing, analysis and understanding. Caused by unconstrained sensing conditions, there exist large intra-class variations inter-class ambiguities realistic videos, which hinder improvement performance for recent vision-based systems. In this paper, we propose a generalized pyramid matching kernel (GPMK) recognizing human actions based on multi-channel “bag words” representation constructed from local...

Datasets collected from the open world unavoidably suffer various forms of randomness or noiseness, leading to ubiquity aleatoric (data) uncertainty. Quantifying such uncertainty is particularly pivotal for object detection, where images contain multi-scale objects with occlusion, obscureness, and even noisy annotations, in contrast centric similar-scale classification. This paper suggests modeling exploiting inherent detection data vision foundation models develops a data-centric reliable...

Natural image matting is one of the major challenges in modern editing, whose goal to estimate fractional opacity foreground layer from an and recover colors foreground/background layers respectively. Most existing methods require complicated mathematics computation are implemented on a single processor. The closed-form [1] relatively faster more robust than previous methods. However, sequential process still time consuming may not produce best results because over-smoothing problems. In...

According to the high-dimensional sparse features of storage textual document, and defects existing in clustering methods which havealready studied by now some other problems, an effective text approach (short for TGSOM-FS-FKM) based on tree-structured growingself-organizing maps (TGSOM) Fuzzy K-Means (FKM) is proposed. It firstly makes preprocess texts, filter majority noisy words usingunsupervised feature selection method. Then it used TGSOM execute first get rough classification initial...

As feature sizes and pitches continue to decrease, more complex correction algorithms are needed solve increasingly difficult geometric configurations. Usage of these results in unacceptably long time-to-mask when applied an entire design. In many cases, the only required a small percentage areas design, not always known prior tapeout. Hotspot fixing (HSF) flows used fix hotspot minimize errors decrease time-to-mask. These involve "recorrecting" using previous output as input HSF flow. This...

The traditional double-threshold endpoint detection method has the phenomenon of missing detection. Therefore, speech recognition (SR) system based on vector quantization (VQ) in this paper proposes an improved algorithm for phenomenon, which effectively avoids problem Then, Mel Frequency Cepstral Coefficients (MFCC) is used to extract characteristic parameters signal, and multistage quantify parameters. Experimental results show that, proposed improves rate text-independent speaker by 8.7%,...