Abstract Artificial intelligence models play a crucial role in monitoring and maintaining railroad infrastructure by analyzing image data of foreign objects on power transmission lines. However, the availability publicly accessible datasets for is limited, rarity anomalies data, combined with restricted sharing, poses challenges training effective object detection models. In this paper, aim to present new dataset lines, evaluating overall performance mainstream context. Taking unique...

Photosynthesis is a key process linking carbon and water cycles, satellite-retrieved solar-induced chlorophyll fluorescence (SIF) can be valuable proxy for photosynthesis. The TROPOspheric Monitoring Instrument (TROPOMI) on the Copernicus Sentinel-5P mission enables significant improvements in providing high spatial temporal resolution SIF observations, but short coverage of data records has limited its applications long-term studies. This study uses machine learning to reconstruct TROPOMI...

Diseases of tomato leaves can seriously damage crop yield and financial rewards. The timely accurate detection diseases is a major challenge in agriculture. Hence, the early diagnosis crucial. emergence deep learning has dramatically helped plant disease detection. However, accuracy models largely depends on quantity quality training data. To solve inter-class imbalance problem improve generalization ability classification model, this paper proposes cycle-consistent...

The tea industry, as one of the most globally important agricultural products, is characterized by pests and diseases that pose a serious threat to yield quality. These often present different scales morphologies, some pest disease target sizes can be tiny difficult detect. To solve these problems, we propose TeaViTNet, multi-scale attention-based detection model combines CNNs Transformers. First, MobileViT used feature extraction backbone network. captures analyzes features in image via...

We present a novel, simple algorithm for mobile robot navigation. Using teach-replay approach, the is manually led along desired path in teaching phase, then autonomously follows that replay phase. The technique requires single off-the-shelf, forward-looking camera with no calibration (including lens distortion). Feature points are automatically detected and tracked throughout image sequence, feature coordinates phase compared those computed previously to determine turning commands robot....

In classification, a large number of features often make it difficult to select appropriate classification features. such situations, feature selection or dimensionality reduction methods play an important role in classification. ReliefF algorithm is one the most successful filtering methods. this paper, some shortcomings are improved, on problem poor stability neighbor samples selection, proposing method using average value multiple random improve anti-volatility algorithm. And redundant...

Improving nitrogen (N) management of small-scale farming systems in developing countries is crucially important for food security and sustainable development world agriculture, but it also very challenging. The N Nutrition Index (NNI) a reliable indicator crop status, there an urgent need to develop effective method non-destructively estimate NNI different smallholder farmer fields guide in-season management. eBee fixed-wing unmanned aerial vehicle (UAV)-based remote sensing system,...

The dynamic interactions between soil, weather and crop management have considerable influences on yield within a region, should be considered in optimizing nitrogen (N) management. objectives of this study were to determine the influence soil type, conditions planting density economic optimal N rate (EONR), evaluate potential benefits site-specific strategies for maize production. experiments conducted two types (black aeolian sandy soils) from 2015 2017, involving different rates (0 300 kg...

The main source of urban waste is the daily life activities residents, and sorting residents’ important for promoting economic recycling, reducing labor costs, protecting environment. However, most residents are unable to make accurate judgments about categories household waste, which severely limits efficiency sorting. We have designed an intelligent bin that enables automatic avoiding extensive knowledge required To ensure waste-classification model high accuracy works in real time,...

The Henan Yellow River Basin is an ecological support belt for the entire basin. It holds a significant position in high-quality development and conservation within Basin. However, due to improper activities, such as urban expansion deforestation of farmland, certain areas region have encountered series issues, posing challenges ecosystem services. scientific foundation sustainable environment established by research on evolution characteristics driving factors service functions. This study...

We present the Binocular Sparse Feature Segmentation (BSFS) algorithm for vision-based person following with a mobile robot. BSFS uses Lucas-Kanade feature detection and matching in order to determine location of image thereby control Matching is performed between two images stereo pair, as well successive video frames. use Random Sample Consensus (RANSAC) scheme segmenting sparse disparity map estimating motion models background. By fusing information, handles difficult situations such...

The growth of vegetation directly maintains the ecological security coal mining areas. It is great significance to monitor dynamic changes in areas and study driving factors spatial division. This focuses on Yima area Henan Province. Utilizing MODIS multi-dimensional explanatory variable data, Theil–Sen Median + Mann–Kendall trend analysis, variation index, Hurst optimal-parameter-based geographical detector model (OPGD) are employed analyze spatiotemporal future trends EVI (enhanced index)...

Land use change is an important factor in atmospheric carbon emissions. Most of the existing studies focus on modeling land pattern for a certain period time future and calculating analyzing However, few have optimized spatial from perspective impact emission constraints structure. Therefore, this study, effects emissions 1990 to 2020 were modeled using flow model Sanmenxia, Henan, China, as example. Then, function under low target was constructed, differential evolution (DE) algorithm used...

China’s urbanization has gradually become the dominant factor in farmland loss, but it remains unclear how impacts supply–demand relationship of agroecosystem services at different rates which turn poses new challenges to ecological security farmland. Taking Huaihe River Basin as a case, this paper analyzes loss from 2000 2020, well changes four types services: food production, socio-economic security, carbon sequestration, and soil conservation, revealing driving mechanisms on sub-regions...

To address the challenge of precise foreign object detection on railway tracks, this study proposes RailSegVITNet, an efficient deep learning-based model that aims to accurately locate tracks and provide necessary information. RailSegVITNet integrates lightweight bottleneck blocks, separable self-attention, feature aggregation balance real-time performance accuracy. It follows encoder–decoder framework, with blocks used for extraction self-attention integrated at different stages enhance...

Water clarity, commonly determined by Secchi disk depth (SDD), is a critical water quality parameter for assessing estuarine ecosystems heath. Since the 1980s, significant changes in SDD have been observed Pearl River Estuary (PRE), robust, economical, and densely populated estuary southern China. However, its long-term patterns associated drivers not yet systematically investigated. In this study, three novel semi-analytical algorithms suitable Landsat 5, 7, 8 images were applied to 36-year...