Weighted Fusion Method Research Articles

Lightweight Detection of Broccoli Heads in Complex Field Environments Based on LBDC-YOLO

Robotically selective broccoli harvesting requires precise lightweight detection models to efficiently detect broccoli heads. Therefore, this study introduces a lightweight and high-precision detection model named LBDC-YOLO (Lightweight Broccoli Detection in Complex Environment—You Look Only Once), based on the improved YOLOv8 (You Look Only Once, Version 8). The model incorporates the Slim-neck design paradigm based on GSConv to reduce computational complexity. Furthermore, Triplet Attention is integrated into the backbone network to capture cross-dimensional interactions between spatial and channel dimensions, enhancing the model’s feature extraction capability under multiple interfering factors. The original neck network structure is replaced with a BiFPN (Bidirectional Feature Pyramid Network), optimizing the cross-layer connection structure, and employing weighted fusion methods for better integration of multi-scale features. The model undergoes training and testing on a dataset constructed in real field conditions, featuring broccoli images under various influencing factors. Experimental results demonstrate that LBDC-YOLO achieves an average detection accuracy of 94.44% for broccoli. Compared to the original YOLOv8n, LBDC-YOLO achieves a 32.1% reduction in computational complexity, a 47.8% decrease in parameters, a 44.4% reduction in model size, and a 0.47 percentage point accuracy improvement. When compared to models such as YOLOv5n, YOLOv5s, and YOLOv7-tiny, LBDC-YOLO exhibits higher detection accuracy and lower computational complexity, presenting clear advantages for broccoli detection tasks in complex field environments. The results of this study provide an accurate and lightweight method for the detection of broccoli heads in complex field environments. This work aims to inspire further research in precision agriculture and to advance knowledge in model-assisted agricultural practices.

Bearing fault diagnosis based on enhanced Canberra distance feature in SDP image

Abstract Feature enhancement is important in mechanical equipment fault diagnosis. A limited set of characteristic parameters is insufficient for diagnosing bearing signals with multiple fault types. The presence of noise increases the difficulty of extracting fault features from images. To address the challenge of diagnosing rolling bearing faults in complex environments, this study presents an enhanced weighted image fusion framework aimed at enhancing fault features within the images, which enables accurate diagnosis of bearing faults using a limited number of features. The proposed method encompasses four distinct stages. In the first stage, a symmetrized dot pattern method is employed to transform one-dimensional time-series data into two-dimensional images, visualizing the signal in a 2D format. In the second stage, image binarization and an improved weighted fusion method are utilized to simplify subsequent processing and enhance the image features. The third stage involves extracting the image’s contrast and maximum singular value to improve the Canberra distance calculation. Finally, the enhanced Canberra distance is used for classifying bearing faults. Performance testing of the image feature enhancement is conducted on various datasets containing rolling bearings. Comparative experiments with alternative enhancement methods demonstrate the superiority of the proposed improved weighted image fusion framework. Comparative experiments with the original Canberra distance validate the effectiveness of the enhanced Canberra distance. Additionally, experiments conducted in noisy environments confirm the robustness of the proposed approach. Furthermore, the image feature enhancement method is applied to other bearing datasets, and the experimental results demonstrate its effectiveness in enhancing fault feature representation and achieving accurate diagnosis of rolling bearings.

Human Fall Detection with Ultra-Wideband Radar and Adaptive Weighted Fusion.

To address the challenges in recognizing various types of falls, which often exhibit high similarity and are difficult to distinguish, this paper proposes a human fall classification system based on the SE-Residual Concatenate Network (SE-RCNet) with adaptive weighted fusion. First, we designed the innovative SE-RCNet network, incorporating SE modules after dense and residual connections to automatically recalibrate feature channel weights and suppress irrelevant features. Subsequently, this network was used to train and classify three types of radar images: time-distance images, time-distance images, and distance-distance images. By adaptively fusing the classification results of these three types of radar images, we achieved higher action recognition accuracy. Experimental results indicate that SE-RCNet achieved F1-scores of 94.0%, 94.3%, and 95.4% for the three radar image types on our self-built dataset. After applying the adaptive weighted fusion method, the F1-score further improved to 98.1%.

Research and application of a novel weight-based evolutionary ensemble model using principal component analysis for wind power prediction

Accurate prediction of wind power is crucial for optimizing wind energy utilization and ensuring the secure, cost-effective, and stable operation of power systems. To address this, a novel approach is proposed that combines a multi-network deep ensemble model with an improved manta ray foraging optimization algorithm (IMRFO) and feature selection for wind power prediction. Firstly, principal component analysis (PCA) is employed to select the principal component with a cumulative contribution rate of 95 %. Secondly, the ensemble prediction model is constructed in the layer-1, incorporating extreme gradient boosting (XGBoost), gated recurrent unit (GRU), and temporal convolutional network (TCN). Additionally, an error-based weighted fusion method is implemented to combine the outputs of three learners. In the layer-2, random forest (RF) employed to make predictions based on the fused results. To further enhance the predictive capabilities of the model, an enhanced MRFO algorithm, incorporating chaos initialization and Gauss-Cauchy mutation strategy, is added. The experiment shows that the deep ensemble model using PCA and IMRFO outperforms XGBoost, GRU, and TCN in predicting. On four datasets, the RMSE of the proposed model is improved by over 45 % compared to them. These findings strongly support the effectiveness and accuracy of the proposed model in this domain.

The Retinex enhancement algorithm for low‐light intensity image based on improved illumination map

AbstractTaken in low‐light intensity conditions, image with low brightness affects processing precision. In this article, the Gamma Function based on the brightness average and weighted fusion method according to gray entropy is proposed, which is combined with the improved Retinex algorithm. First, the maximum values of R, G, and B channels in original image are extracted to generate the primary illumination map. Second, the illumination map is optimized and adjusted via the Gamma correction function based on the average brightness value. Finally, the illumination map and detail layer are fused by a weighted fusion algorithm of gray entropy to obtain the reflection map. Reflection maps are used as enhancement. The algorithm proposed in this article can improve the brightness and maintain light distribution in the original image with higher precision and less color distortion.

Tactile texture recognition of multi-modal bionic finger based on multi-modal CBAM-CNN interpretable method

To solve the problem of lacking surface information while using a single-modal sensor to scan fine textures, a multi-modal bionic finger tactile sensor is used to perceive the information contained in textures, and a multi-modal tactile texture recognition method is proposed to address the issue of poor interpretability for tactile texture recognition models. First, the experimental platform of multi-modal bionic finger tactile texture information acquisition system is built, which can simultaneously collect 14 channels of tactile electrode signals, one channel of pressure signals and one channel of temperature signals. Second, the multi-modal tactile signals can be fused at the data level after signal processing, and then the wavelet transform method is carried out on the basis of data fusion. Third, a multi-modal CBAM-CNN network (Convolutional Block Attention Module, Convolutional Neural Network) is established to extract the multi-modal tactile features after the wavelet transform and the multi-modal tactile features are fused at feature level. To verify the performance of multi-modal CBAM-CNN model, 27(3*9) kinds of millimeter-level fine textures of 3 datasets are tested and the recognition accuracy of each dataset is 98.47%, 97.63%, 99.16%. And it can also be seen that the 14-channel modal data contributes the most to the performance of multi-modal CBAM-CNN model for various textures. The results of ablation experiments prove that the identification accuracy of the fused multi-modal CBAM-CNN is higher than the single-modal CBAM-CNN no matter in time domain or time–frequency domain. Moreover, the accuracy of the weighted fusion method based on correlation function is higher than horizontal fusion method in 14 channels of tactile electrode signals modal. Finally, to interpret the classification decision of the multi-modal CBAM-CNN network in each modal intuitively, the improved multi-modal CAM (Class Activation Map) algorithm is adopted to visualize the interest regions of the samples in detail, and the activated areas in the map make the identification result of CBAM-CNN in each modal more transparent and interpretable.

Weighted information index mining of key nodes through the perspective of evidential distance

Many strategies have been developed to mine critical nodes in the networks, however, most of them are based solely on the network topology, with no regard for the diversity of information between nodes. Using a weighted fusion method, we propose a method that considers neighboring nodes of second order and their similarity in this paper. Based on the evidence distance, an information distance Id is computed to find out how similar nodes are, and second-order adjacent nodes are used to reflect node information. Information distance is used to determine the weight of a node, and a weighted information index (WII) is used to evaluate how important a node is by combining the weight and information of nodes. The proposed method is tested by identifying influential nodes in twelve real-world networks and comparing their performance with the other seven methods. The Susceptible–Infected (SI) model and two evaluation indices have been applied to twelve real-world networks to evaluate the feasibility of our method. The findings indicate that the WII exhibits the most effective differentiation, thereby demonstrating the efficacy and practicality of our approach.

Optimizing Controls to Track Moving Targets in an Intelligent Electro-Optical Detection System

Electro-optical detection systems face numerous challenges due to the complexity and difficulty of targeting controls for “low, slow and tiny” moving targets. In this paper, we present an optimal model of an advanced n-step adaptive Kalman filter and gyroscope short-term integration weighting fusion (nKF-Gyro) method with targeting control. A method is put forward to improve the model by adding a spherical coordinate system to design an adaptive Kalman filter to estimate target movements. The targeting error formation is analyzed in detail to reveal the relationship between tracking controller feedback and line-of-sight position correction. Based on the establishment of a targeting control coordinate system for tracking moving targets, a dual closed-loop composite optimization control model is proposed. The outer loop is used for estimating the motion parameters and predicting the future encounter point, while the inner loop is used for compensating the targeting error of various elements in the firing trajectory. Finally, the modeling method is substituted into the disturbance simulation verification, which can monitor and compensate for the targeting error of moving targets in real time. The results show that in the optimal model incorporating the nKF-Gyro method with targeting control, the error suppression was increased by up to 36.8% compared to that of traditional KF method and was 25% better than that of the traditional nKF method.

A regional digital bathymetric model fusion method based on topographic slope: A case study of the south China sea and surrounding waters

High-resolution seafloor topography is important in scientific research and marine engineering in regard to marine resource development and environmental protection monitoring. In this study, multi-dimensional comparisons were made between GEBCO_2022, SRTM15_V2.5.5, SRTM30_PLUS, SYNBATH_V1.0, ETOPO_2022, and topo_25.1 in the South China Sea and surrounding waters (SCS). This study has found that ETOPO_2022 had the best overall accuracy and reliability. Based on the results of the model accuracy analysis and by considering the topographic slope, ETOPO_2022, GEBCO_2022, and SRTM15_V2.5.5 were weighted and fused to form a fusion model. The error of the fusion model was 94.80% concentrated in (−100–100 m). When compared with GEBCO_2022, SRTM15_V2.5.5, SRTM30_PLUS, SYNBATH_V1.0, ETOPO_2022, and topo_25.1, the RMSE was reduced by 2%, 9%, 62%, 15%, 1%, and 73%, respectively. The slope-based weighted fusion method has been shown that it can overcome the limitations of a single data source and provide a reference for timely reconstruction and updating of large-scale seafloor topography.

A New Progressive Multisource Domain Adaptation Network With Weighted Decision Fusion.

Multisource unsupervised domain adaptation (MUDA) is an important and challenging topic for target classification with the assistance of labeled data in source domains. When we have several labeled source domains, it is difficult to map all source domains and target domain into a common feature space for classifying the targets well. In this article, a new progressive multisource domain adaptation network (PMSDAN) is proposed to further improve the classification performance. PMSDAN mainly consists of two steps for distribution alignment. First, the multiple source domains are integrated as one auxiliary domain to match the distribution with the target domain. By doing this, we can generally reduce the distribution discrepancy between each source and target domains, as well as the discrepancy between different source domains. It can efficiently explore useful knowledge from the integrated source domain. Second, to mine assistance knowledge from each source domain as much as possible, the distribution of the target domain is separately aligned with that of each source domain. A weighted fusion method is employed to combine the multiple classification results for making the final decision. In the optimization of domain adaption, weighted hybrid maximum mean discrepancy (WHMMD) is proposed, and it considers both the interclass and intraclass discrepancies. The effectiveness of the proposed PMSDAN is demonstrated in the experiments comparing with some state-of-the-art methods.

Joint QoS Prediction for Web Services based on Deep Fusion of Features

In order to solve the problem of insufficient accuracy of Web service QoS prediction, a joint QoS prediction method for Web services based on the deep fusion of features was proposed by considering the hidden environmental preference information in QoS and the common characteristics of multi-class QoS. QoS data was modeled as user-service bipartite graph at first, then, multi-component graph convolution neural network was used for feature extraction and mapping, and weighted fusion method was used for the same dimensional mapping of multi-class of QoS features. Subsequently, the attention factor decomposition machine was used to extract the first-order features, second-order interactive features and high-order interactive features of the mapped feature vector. Finally, the results of each part were combined to achieve the joint QoS prediction. The experimental results show that the proposed method is superior to the existing QoS prediction methods in terms of root mean square error (RMSE) and average absolute error (MAE).

Zonotopic Distributed Fusion Over Binary Sensor Networks With Bit Rate Allocation: A Coding-Decoding Approach.

In this article, the zonotopic distributed fusion estimation problem is investigated for a class of general nonlinear systems over binary sensor networks subject to unknown-but-bounded (UBB) noises. The network communication from nodes to the fusion center is confined to the limited bit rate. To alleviate the impact from less measurement information of the binary sensor, a modified innovation is constructed to improve the estimation accuracy. Then, a novel coding-decoding approach is proposed to ensure that the decoder has the ability to decode information from each node. Based on the matrix weighting fusion method, a distributed fusion algorithm is put forward under the zonotopic set-membership filtering framework, and the F -radius of the local zonotopic sets are derived and minimized by selecting the filtering gain parameters. Moreover, the bit rate allocation scheme and the weighting coefficients are determined by resolving two optimization problems. In addition, a sufficient condition is established to guarantee the uniform boundedness of the F -radius of the fused zonopotic. Finally, the ballistic object tracking systems is utilized to illustrate the availability of the presented algorithm.

An improved stacking-based model for wave height prediction

<p>Wave height prediction is hampered by the volatility and unpredictability of ocean data. Traditional single predictors are inadequate in capturing this complexity, and weighted fusion methods fail to consider inter-model correlations, resulting in suboptimal performance. To overcome these challenges, we presented an improved stacking-based model that combined the long short-term memory (LSTM) network with extremely randomized trees (ET) for wave height prediction. Initially, features with weak correlation to wave height were excluded using the Pearson correlation coefficient. Subsequently, a stacking ensemble tailored for time series cross-validation was deployed, employing LSTM and ET as base learners to capture temporal and feature-specific patterns, respectively. Lasso regression was utilized as the meta-learner, harmonizing these insights to improve accuracy by leveraging the strengths of each model across different dimensions of the data. Validation using datasets from four buoy stations demonstrated the superior predictive capability of our proposed model over single predictors such as temporal convolutional networks (TCN) and XGBoost, and fusion methods like LSTM-ET-BP.</p>

A Parallel Multimodal Integration Framework and Application for Cake Shopping

Multimodal interaction systems can provide users with natural and compelling interactive experiences. Despite the availability of various sensing devices, only some commercial multimodal applications are available. One reason may be the need for a more efficient framework for fusing heterogeneous data and addressing resource pressure. This paper presents a parallel multimodal integration framework that ensures that the errors and external damages of integrated devices remain uncorrelated. The proposed relative weighted fusion method and modality delay strategy process the heterogeneous data at the decision level. The parallel modality operation flow allows each device to operate across multiple terminals, reducing resource demands on a single computer. The universal fusion methods and independent devices further remove constraints on the integrated modality number, providing the framework with extensibility. Based on the framework, we develop a multimodal virtual shopping system, integrating five input modalities and three output modalities. The objective experiments show that the system can accurately fuse heterogeneous data and understand interaction intent. User studies indicate the immersive and entertaining of multimodal shopping. Our framework proposes a development paradigm for multimodal systems, fostering multimodal applications across various domains.

Bed Separation Formation Mechanism and Water Inrush Evaluation in Coal Seam Mining under a Karst Cave Landform

Understanding the formation mechanism of bed separation in coal seam mining under a karst landform is needed for the prevention and control of roof-separated water damage in such areas. This research used a mine in the northern Guizhou coalfield, China, as a case study, and applied theoretical analysis, numerical simulation, and on-site measurement to develop a circular cave structure model in a key stratum. The dynamic evolution of a separation bed was analyzed from several aspects, including the formation mechanism, development location, the mechanical condition of local karst caves, fracture evolution, and fractal rules. Verification using in situ measurements is presented for the case study mine, and a quantitative evaluation method for water inrush from bed separation and improved fusion weighting is proposed based on a cloud model. The research results indicate the following: (1) Tensile cracks are prone to occur above and below a karst cave, which produce an impact of connectivity on the separated space. (2) When the working face advances to 270 m in coal mining, longitudinal tensile cracks below the karst cave gradually increase and the width of the bed separation crack shrinks to 68.2 m, with a maximum separation layer height of 3.01 m. (3) Based on the cloud model and the improved weighted fusion method, the risk of water inrush in bed separation is judged as “high”. The En of the cloud digital features is 0.0622 and the He is 0.0307, achieving a quantitative evaluation of water inrush in the separation layer that is consistent with on-site practice, and is highly stable and reliable. This study improves the understanding of the development pattern of bed separation and water inrush risk assessment in coal seam mining under a karst cave landform.

Fusion of Environmental Sensors for Occupancy Detection in a Real Construction Site

Internet-of-Things systems are increasingly being installed in buildings to transform them into smart ones and to assist in the transition to a greener future. A common feature of smart buildings, whether commercial or residential, is environmental sensing that provides information about temperature, dust, and the general air quality of indoor spaces, assisting in achieving energy efficiency. Environmental sensors though, especially when combined, can also be used to detect occupancy in a space and to increase security and safety. The most popular methods for the combination of environmental sensor measurements are concatenation and neural networks that can conduct fusion in different levels. This work presents an evaluation of the performance of multiple late fusion methods in detecting occupancy from environmental sensors installed in a building during its construction and provides a comparison of the late fusion approaches with early fusion followed by ensemble classifiers. A novel weighted fusion method, suitable for imbalanced samples, is also tested. The data collected from the environmental sensors are provided as a public dataset.

Spatial Domain Image Fusion with Particle Swarm Optimization and Lightweight AlexNet for Robotic Fish Sensor Fault Diagnosis.

Safety and reliability are vital for robotic fish, which can be improved through fault diagnosis. In this study, a method for diagnosing sensor faults is proposed, which involves using Gramian angular field fusion with particle swarm optimization and lightweight AlexNet. Initially, one-dimensional time series sensor signals are converted into two-dimensional images using the Gramian angular field method with sliding window augmentation. Next, weighted fusion methods are employed to combine Gramian angular summation field images and Gramian angular difference field images, allowing for the full utilization of image information. Subsequently, a lightweight AlexNet is developed to extract features and classify fused images for fault diagnosis with fewer parameters and a shorter running time. To improve diagnosis accuracy, the particle swarm optimization algorithm is used to optimize the weighted fusion coefficient. The results indicate that the proposed method achieves a fault diagnosis accuracy of 99.72% when the weighted fusion coefficient is 0.276. These findings demonstrate the effectiveness of the proposed method for diagnosing depth sensor faults in robotic fish.

Wide-Area Composite Load Parameter Identification Based on Multi-Residual Deep Neural Network.

Accurate and practical load modeling plays a critical role in the power system studies including stability, control, and protection. Recently, wide-area measurement systems (WAMSs) are utilized to model the static and dynamic behavior of the load consumption pattern in real-time, simultaneously. In this article, a WAMS-based load modeling method is established based on a multi-residual deep learning structure. To do so, a comprehensive and efficient load model founded on combination of impedance-current-power and induction motor (IM) is constructed at the first step. Then, a deep learning-based framework is developed to understand the time-varying and complex behavior of the composite load model (CLM). To do so, a residual convolutional neural network (ResCNN) is developed to capture the spatial features of the load at different location of the large-scale power system. Then, gated recurrent unit (GRU) is used to fully understand the temporal features from highly variant time-domain signals. It is essential to provide a balance between fast and slow variant parameters. Thus, the designed structure is implemented in a parallel manner to fulfill the balance and moreover, weighted fusion method is used to estimate the parameters, as well. Consequently, an error-based loss function is reformulated to improve the training process as well as robustness in the noisy conditions. The numerical experiments on IEEE 68-bus and Iranian 95-bus systems verify the effectiveness and robustness of the proposed load modeling approach. Furthermore, a comparative study with some relevant methods demonstrates the superiority of the proposed structure. The obtained results in the worst-case scenario show error lower than 0.055% considering noisy condition and at least 50% improvement comparing the several state-of-art methods.

TASTNet: An end-to-end deep fingerprinting net with two-dimensional attention mechanism and spatio-temporal weighted fusion for video content authentication

In this paper, a robust fingerprinting scheme for video content authentication with two-dimensional attention mechanism and spatio-temporal weighted fusion called TASTNet is proposed, which can automatically extracts key spatio-temporal features from the input video and maps them to the corresponding fingerprint. Detailedly, the two-dimensional attention mechanism is applied to resist different kinds of digital manipulations for robustness enhancement. To incorporate perceptual characteristics, a spatio-temporal weighted fusion method based on LTSM is presented to integrate frame-level features into video-level features while retaining the temporal order. In the process of fusion, key frames are allocated with larger weights according to inter-frame correlation. With these two steps, we can obtain representative video features that contain principal perception information. In addition, the proposed scheme utilizes deep metric learning for training, and we design multiple constraints to make the generated fingerprint more compact and discriminable. Extensive experiments demonstrate that our scheme can achieve superior performances with respect to robustness and discrimination compared with some state-of-the-art schemes.

Multi‐sensor data fusion for swine house environment evaluation and control

AbstractData fusion from various sensors can obtain more accurate and reliable results than a single monitoring sensor. The data of air quality in the swine breeding environment comes from various sensors; therefore, the data must be fused. In this study, statistical criteria, batch estimation theory, a fuzzy comprehensive evaluation (FCE) method, as well as an improved analytic hierarchy process (IAHP), were combined to provide accurate data for multi‐sensor data fusion modeling. To make a fair comparison, four statistical criteria were employed to process abnormal data gross errors. Besides, an improved batch estimation adaptive weighted fusion method was adopted to, respectively, integrate the data from temperature, humidity, and concentrations of ammonia (NH3), carbon dioxide (CO2), and hydrogen sulfide (H2S) of air quality at the data level. The predetermined weights were given using IAHP combined with the knowledge of experts. To decide the fuzzy membership values of each environmental parameter, we presented a method to establish fuzzy membership functions and devised a calculation method for related parameters. The gaussmf type membership function was employed to build a global fusion model that fits well with IAHP–FCE. The IAHP–FCE method was exploited to comprehensively integrate multi‐sensor data at the decision level. Experimental results show that the presented method can improve the data accuracy and provide decision results and scientific basis for the precision control of swine breeding environment.