Search Results (79)

Accurate and rapid aquatic organism recognition is a core technology for fisheries automation and aquatic organism statistical research. However, due to absorption and scattering effects, images of aquatic organisms often suffer from poor contrast and color distortion. Additionally, the clustering behavior of aquatic organisms often leads to occlusion, further complicating the identification task. This study proposes a lightweight object detection model, Mobile-YOLO, for the recognition of four representative aquatic organisms, namely holothurian, echinus, scallop, and starfish. Our model first utilizes the Mobile-Nano backbone network we proposed, which enhances feature perception while maintaining a lightweight design. Then, we propose a lightweight detection head, LDtect, which achieves a balance between lightweight structure and high accuracy. Additionally, we introduce Dysample (dynamic sampling) and HWD (Haar wavelet downsampling) modules, aiming to optimize the feature fusion structure and achieve lightweight goals by improving the processes of upsampling and downsampling. These modules also help compensate for the accuracy loss caused by the lightweight design of LDtect. Compared to the baseline model, our model reduces Params (parameters) by 32.2%, FLOPs (floating point operations) by 28.4%, and weights (model storage size) by 30.8%, while improving FPS (frames per second) by 95.2%. The improvement in mAP (mean average precision) can also lead to better accuracy in practical applications, such as marine species monitoring, conservation efforts, and biodiversity assessment. Furthermore, the model’s accuracy is enhanced, with the mAP increased by 1.6%, demonstrating the advanced nature of our approach. Compared with YOLO (You Only Look Once) series (YOLOv5-12), SSD (Single Shot MultiBox Detector), EfficientDet (Efficient Detection), RetinaNet, and RT-DETR (Real-Time Detection Transformer), our model achieves leading comprehensive performance in terms of both accuracy and lightweight design. The results indicate that our research provides technological support for precise and rapid aquatic organism recognition. Full article

Haze caused by atmospheric scattering often leads to color distortion, reduced contrast, and diminished clarity, which significantly degrade the quality of remote sensing images. To address these issues, we propose a novel network called DWTMA-Net that integrates discrete wavelet transform with multi-dimensional attention, aiming to restore image information in both the frequency and spatial domains to enhance overall image quality. Specifically, we design a wavelet transform-based downsampling module that effectively fuses frequency and spatial features. The input first passes through a discrete wavelet block to extract frequency-domain information. These features are then fed into a multi-dimensional attention block, which incorporates pixel attention, Fourier frequency-domain attention, and channel attention. This combination allows the network to capture both global and local characteristics while enhancing deep feature representations through dimensional expansion, thereby improving spatial-domain feature extraction. Experimental results on the SateHaze1k, HRSD, and HazyDet datasets demonstrate the effectiveness of the proposed method in handling remote sensing images with varying haze levels and drone-view scenarios. By recovering both frequency and spatial details, our model achieves significant improvements in dehazing performance compared to existing state-of-the-art approaches. Full article

Partial discharge (PD) analysis is critical for diagnosing insulation degradation in high-voltage equipment. While conventional methods struggle with multi-source PD classification due to signal overlap and noise, this study proposes a hybrid approach combining five time–frequency analysis (TFA) techniques with deep learning (GoogLeNet for simulation, ResNet50 for experiments). PD data are generated through Finite Element Method (FEM) simulations and validated via laboratory experiments. The Scatter Wavelet Transform (SWT) achieves 96.67% accuracy (F1-score: 0.967) in simulation and perfect 100% accuracy (F1-score: 1.000) in experiments, outperforming other TFAs like HHT (70.00% experimental accuracy). The Wigner–Ville Distribution (WVD) also shows strong experimental performance (94.74% accuracy, AUC: 0.947), though its computational complexity limits real-time use. These results demonstrate the SWT’s superiority in handling real-world noise and multi-source PD signals, providing a robust framework for insulation diagnostics in power systems. Full article

The leaf water content (LWC) of citrus is a pivotal indicator for assessing citrus water status. Addressing the limitations of traditional hyperspectral modelling methods, which rely on single preprocessing techniques and struggle to fully exploit the complex information within spectra, this study proposes a novel strategy for estimating citrus LWC by integrating spectral preprocessing combinations with an enhanced deep learning architecture. Utilizing a citrus plantation in Guangxi as the experimental site, 240 leaf samples were collected. Seven preprocessing combinations were constructed based on multiplicative scatter correction (MSC), continuous wavelet transform (CWT), and first derivative (1st D), and a new multichannel network, EDPNet (Ensemble Data Preprocessing Network), was designed for modelling. Furthermore, this study incorporated an attention mechanism within EDPNet, comparing the applicability of SE Block, SAM, and CBAM in integrating spectral combination information. The experiments demonstrated that (1) the triple preprocessing combination (MSC + CWT + 1st D) significantly enhanced model performance, with the prediction set R² reaching 0.8036, a 13.86% improvement over single preprocessing methods, and the RMSE reduced to 2.3835; (2) EDPNet, through its multichannel parallel convolution and shallow structure design, avoids excessive network depth while effectively enhancing predictive performance, achieving a prediction accuracy (R² = 0.8036) that was 5.58–9.21% higher than that of AlexNet, VGGNet, and LeNet-5, with the RMSE reduced by 9.35–14.65%; and (3) the introduction of the hybrid attention mechanism CBAM further optimized feature weight allocation, increasing the model’s R² to 0.8430 and reducing the RMSE to 2.1311, with accuracy improvements of 2.08–2.36% over other attention modules (SE, SAM). This study provides a highly efficient and accurate new method for monitoring citrus water content, offering practical significance for intelligent orchard management and optimal resource allocation. Full article

Underwater image enhancement (UIE) is inherently challenging due to complex degradation effects such as light absorption and scattering, which result in color distortion and a loss of fine details. Most existing methods focus on spatial-domain processing, often neglecting the frequency-domain characteristics that are crucial for effectively restoring textures and edges. In this paper, we propose a novel UIE framework, the Wavelet-based Enhancement Diffusion Model (WEDM), which integrates frequency-domain decomposition with diffusion models. The WEDM consists of two main modules: the Wavelet Color Compensation Module (WCCM) for color correction in the LAB space using discrete wavelet transform, and the Wavelet Diffusion Module (WDM), which replaces traditional convolutions with wavelet-based operations to preserve multi-scale frequency features. By combining residual denoising diffusion with frequency-specific processing, the WEDM effectively reduces noise amplification and high-frequency blurring. Ablation studies further demonstrate the essential roles of the WCCM and WDM in improving color fidelity and texture details. Our framework offers a robust solution for underwater visual tasks, with promising applications in marine exploration and ecological monitoring. Full article

Accurate operational methods used to measure, verify, and report changes in biomass at large spatial scales are required to support conservation initiatives. In this study, we demonstrate that machine learning can be used to model aboveground biomass (AGB) in both tropical and temperate forest ecosystems when provided with a sufficiently large training dataset. Using wavelet-transformed airborne hyperspectral imagery, we trained a shallow neural network (SNN) to model AGB. An existing global AGB map developed as part of the European Space Agency’s DUE GlobBiomass project served as the training data for all study sites. At the temperate site, we also trained the model on airborne-LiDAR-derived AGB. In comparison, for all study sites, we also trained a separate deep convolutional neural network (3D-CNN) with the hyperspectral imagery. Our results show that extracting both spatial and spectral features with the 3D-CNN produced the lowest RMSE across all study sites. For example, at the tropical forest site the Tortuguero conservation area, with the 3D-CNN, an RMSE of 21.12 Mg/ha (R² of 0.94) was reached in comparison to the SNN model, which had an RMSE of 43.47 Mg/ha (R² 0.72), accounting for a ~50% reduction in prediction uncertainty. The 3D-CNN models developed for the other tropical and temperate sites produced similar results, with a range in RMSE of 13.5 Mg/ha–31.18 Mg/ha. In the future, as sufficiently large field-based datasets become available (e.g., the national forest inventory), a 3D-CNN approach could help to reduce the uncertainty between hyperspectral reflectance and forest biomass estimates across tropical and temperate bioclimatic domains. Full article

Bearing condition monitoring and prognosis are crucial tasks for ensuring the proper operation of rotating machinery and mechanical systems. Vibration signal analysis is one of the most effective techniques for bearing condition monitoring and prognosis. In this study, the wavelet scattering transform, derived from wavelet transforms and incorporating concepts from scattering transform and convolutional network architectures, was utilized to extract quantitative features from vibration signals. The number of wavelet scattering coefficients obtained from vibration signals of different lengths varied due to the use of predefined wavelet and scaling filters in the wavelet scattering network. Additionally, these wavelet scattering coefficients are associated with different scattering paths within the corresponding wavelet scattering networks. Eight different lengths of vibration signals, associated with fifteen classes of rolling element bearing faults and conditions, were investigated using wavelet scattering feature extraction. The classes of rolling element bearing faults and conditions included normal bearings as well as ball and inner race faults with various fault diameters ranging from 0.007 inches to 0.028 inches. For the specific dataset validated, the computational results indicated that excellent bearing fault classification was achievable using wavelet scattering feature vectors derived from vibration signals with lengths of at least 6000 samples. Full article

Accurate identification and monitoring of aerosol properties is crucial for understanding their sources and impacts on human health and the environment. Therefore, we propose a feature extraction and attribute recognition method from in situ light-scattering measurements based on Bayesian Optimization, wavelet scattering transform, and long short-term memory neural network (BO-WST-LSTM), with empirical mode decomposition (EMD) and independent component analysis (ICA) algorithm for signal preprocessing. In this study, an experimental platform was utilized to gather light-scattering signals from particles with varying characteristics. The signals are then processed using the EMD-ICA noise reduction technique. Then, the wavelet scattering network is used to realize the adaptive extraction of the characteristics of the particle light-scattering signal, and the Bayesian Optimization model is used to optimize the hyperparameters of the LSTM neural network. The extracted scattering coefficient matrix is input into the LSTM for iterative training. Finally, the SoftMax layer’s probability classification method is applied to the identification of particle attributes. The results show that the multi-angle particle light-scattering signal detection system designed and built in this study performs well and is capable of effectively collecting particle light-scattering signals. At the same time, the proposed new method for particle property recognition demonstrates good classification performance for six different types of particles with a particle size of 2 μm, achieving a classification accuracy of 98.83%. This proves its effectiveness in recognizing particle properties and provides a solid foundation for particle identification. Full article

Heavy metal pollution in agricultural land poses significant threats to both the ecological environment and human health. Therefore, the rapid and accurate prediction of heavy metal content in agricultural soil is crucial for environmental protection and soil remediation. Acknowledging the limitations of traditional single linear or nonlinear machine learning models in terms of prediction accuracy, this study developed an ensemble learning model that integrates multiple linear or nonlinear learning models with a random forest (RF) model to improve both the prediction accuracy and reliability. In this study, we selected a typical copper (Cu) polluted area in the Pearl River Delta of Guangdong Province as the research site and collected Cu content data and indoor soil reflectance spectral data from 269 surface soil samples. First, the soil spectral data were preprocessed using Savitzky–Golay (SG) smoothing, multiplicative scattering correction (MSC), and continuous wavelet transform (CWT) to reduce noise interference. Next, principal components analysis (PCA) was employed to reduce the dimensionality of the preprocessed spectral data, eliminating redundant features and lowering the computational complexity. Finally, based on the dimensionality-reduced data and Cu content, we established a stacked ensemble learning model, where the base models included SVR, PLSR, BPNN, and XGBoost, with RF serving as the meta-model to estimate the soil heavy metal content. To evaluate the performance of the stacking model, we compared its prediction accuracy with that of individual models. The results indicate that, compared to the traditional machine learning models, the prediction accuracy of the stacking model was superior (R² = 0.77; RMSE = 7.65 mg/kg; RPD = 2.29). This suggests that the integrated algorithm demonstrates a greater robustness and generalization capability. This study presents a method to improve soil heavy metal content estimation using hyperspectral technology, ensuring a robust model that supports policymakers in making informed decisions about land use, agriculture, and environmental protection. Full article

Sonar automatic target recognition (ATR) systems suffer from complex acoustic scattering, background clutter, and waveguide effects that are ever-present in the ocean. Traditional signal processing techniques often struggle to distinguish targets when noise and complicated target geometries are introduced. Recent advancements in machine learning and wavelet theory offer promising directions for extracting informative features from sonar return data. This work introduces a feature extraction and dimensionality reduction technique using the invariant scattering transform and Sparse Multi-view Discriminant Analysis for identifying highly informative features in the PONDEX09/PONDEX10 datasets. The extracted features are used to train a support vector machine classifier that achieves an average classification accuracy of 97.3% using six unique targets. Full article

The study of chlorophyll fluorescence parameters is very important for understanding plant photosynthesis. Monitoring cotton chlorophyll fluorescence parameters via spectral technology can aid in understanding the photosynthesis, growth, and stress of cotton fields in real time and provide support for cotton growth regulation and planting management. In this study, cotton plot experiments with different water treatments were set up to obtain the spectral reflectance of the cotton canopy, the maximum photochemical quantum yield (Fv/Fm), and the photochemical quenching coefficient (qP) of leaves at different growth stages. Support vector machine regression (SVR), random forest regression (RFR), and artificial neural network regression (ANNR) were used to establish a fluorescence parameter inversion model of the cotton canopy leaves. The results show that the original spectrum was transformed by multivariate scattering correction (MSC), the standard normal variable (SNV), and continuous wavelet transform (CWT), and the model constructed with Fv/Fm passed accuracy verification. The SNV-SVR model at the budding stage, the MSC-SVR model at the early flowering stage, the SNV-SVR model at the full flowering stage, the MSC-SVR model at the flowering stage, and the CWT-SVR model at the full boll stage had the highest estimation accuracy. The accuracies of the three spectral preprocessing and qP models were verified, and the MSC-SVR model at the budding stage, SNV-SVR model at the early flowering stage, MSC-SVR model at the full flowering stage, SNV-SVR model at the flowering stage, and CWT-SVR model at the full boll stage presented the highest estimation accuracies. Full article

Water scarcity is one of the most significant environmental factors that inhibits photosynthesis and decreases the growth and productivity of plants. Using the deep learning convolutional neural network (CNN) model, this study evaluates the ability of spectroscopy to estimate leaf water content (LWC) in fruit trees. During midday, spectral data were acquired from leaf samples obtained from three distinct varieties of fruit trees, encompassing the spectral range spanning from 350 to 2500 nm. Then, for spectral preprocessing, the fractional order derivative (FOD) and continuous wavelet transform (CWT) algorithms were used to reduce the effects of scattering and noise on the collected spectra. Finally, the CNN model was developed to predict LWC in different fruit trees. The results showed that: (1) The spectra treated with CWT and FOD could improve the spectrum expression ability by improving the correlation between spectra and LWC. The correlation level of FOD treatment was higher than that of CWT treatment. (2) The CNN model was developed using FOD _1.2, and CWT ₃ performed better than other traditional machine learning methods, such as RFR, SVR, and PLSR. (3) Further validation using additional samples demonstrated that the CNN model had good stability and quantitative prediction capability for the LWC of fruit trees (R² > 0.95, root mean square error (RMSE) < 1.773%, and relative percentage difference (RPD) > 4.26). The results may provide an effective way to predict fruit LWC using a CNN-based model. Full article

Roads play a crucial role in urban transportation by facilitating the movement of materials within a city. The condition of road surfaces, such as damage and road facilities, directly affects traffic flow and influences decisions related to urban transportation maintenance and planning. To gather this information, we propose the Detecting and Clustering Framework for sensing road surface conditions based on crowd-sourced trajectories, utilizing various sensors (GPS, orientation sensors, and accelerometers) found in smartphones. Initially, smartphones are placed randomly during users’ travels on the road to record the road surface conditions. Then, spatial transformations are applied to the accelerometer data based on attitude readings, and heading angles are computed to store movement information. Next, the feature encoding process operates on spatially adjusted accelerations using the wavelet scattering transformation. The resulting encoding results are then input into the designed LSTM neural network to extract bump features of the road surface (BFRSs). Finally, the BFRSs are represented and integrated using the proposed two-stage clustering method, considering distances and directions. Additionally, this procedure is also applied to crowd-sourced trajectories, and the road surface condition is computed and visualized on a map. Moreover, this method can provide valuable insights for urban road maintenance and planning, with significant practical applications. Full article

Single-image dehazing is a pivotal challenge in computer vision that seeks to remove haze from images and restore clean background details. Recognizing the limitations of traditional physical model-based methods and the inefficiencies of current attention-based solutions, we propose a new dehazing network combining an innovative Haze-Aware Attention Module (HAAM) with a Multiscale Frequency Enhancement Module (MFEM). The HAAM is inspired by the atmospheric scattering model, thus skillfully integrating physical principles into high-dimensional features for targeted dehazing. It picks up on latent features during the image restoration process, which gives a significant boost to the metrics, while the MFEM efficiently enhances high-frequency details, thus sidestepping wavelet or Fourier transform complexities. It employs multiscale fields to extract and emphasize key frequency components with minimal parameter overhead. Integrated into a simple U-Net framework, our Haze-Aware Attention Network (HAA-Net) for single-image dehazing significantly outperforms existing attention-based and transformer models in efficiency and effectiveness. Tested across various public datasets, the HAA-Net sets new performance benchmarks. Our work not only advances the field of image dehazing but also offers insights into the design of attention mechanisms for broader applications in computer vision. Full article

Chipless RFID tags have attractive low-cost advantages. However, traditional RFID anti-collision algorithms cannot be applied due to a lack of computing and processing capabilities. Problems with multitag detection must be solved to commercialize chipless RFID tags. In this paper, an innovative method for frequency-domain chipless RFID tag detection is proposed. The tags’ scattered signals are processed via wavelet analysis, and a time–frequency plot that can read the code is obtained. When the distance between tags is too close to distinguish in the time–frequency plot, independent component analysis is used to separate individual scattered signals from mixed echo signals; then, the code is read by means of wavelet analysis. To validate the proposed method, C-shaped frequency-domain chipless RFID tag models and a multitag detection simulation scenario were constructed in selected software. The short-time matrix pencil method (STMPM), short-time Fourier transform (STFT), and the proposed method were compared. When the tag spacing is 0.05 m, the code can be read successfully. Compared with the STMPM, the proposed method greatly reduces the computational quantity and shortens the reading time. Furthermore, adjustment of the window width and search step parameters is avoided. Full article