Fusion Algorithm Research Articles

Multi-Scale Candidate Fusion and Optimization-Based 3D Object Detection Algorithm

To address the issues of target omission and the inclusion of a large number of background points in keypoint sampling for point cloud-based object detection, an improved algorithm based on the PV-RCNN network is introduced. This approach employs both a regional proposal fusion network and weighted Non-Maximum Suppression (NMS) to merge proposals generated at various scales while eliminating redundancy. A segmentation network is utilized to segment foreground points from the original point cloud, and object center points are identified based on these proposals. Gaussian density functions are employed for regional density estimation, which assigns different sampling weights to solve the problem of difficult sampling in sparse areas. Experimental evaluations on the KITTI dataset indicate that the algorithm enhances the average precision at medium difficulty levels by 0.39%, 1.31%, and 0.63%for cars, pedestrians, and cyclists, respectively. Generalization experiments were also conducted on the Waymo dataset. The results suggest that the introduced algorithm achieves higher accuracy compared to most of the existing 3D object detection networks.

Research on AGV Path Planning Based on Improved Directed Weighted Graph Theory and ROS Fusion

This article addresses the common issues of insufficient computing power and path congestion for automated guided vehicles (AGVs) in real-world production environments, as well as the shortcomings of traditional path-planning algorithms that mainly consider the shortest path while ignoring vehicle turning time and stability. We propose a secondary path-planning method based on an improved directed weighted graph theory integrated with an ROS. Firstly, the production environment is modeled in detail to identify the initial position of the AGV. Secondly, the operational area is systematically divided, key nodes are selected and optimized, and a directed weighted graph is constructed with optimized weights. It is integrated with the ROS for path planning, using the Floyd algorithm to find the optimal path. The effectiveness and superiority of this method have been demonstrated through simulation verification and actual AGV operation testing. The path planning strategy and fusion algorithm proposed in this article that comprehensively considers distance and angle steering are simple and practical, effectively reducing production costs for enterprises. This method is suitable for logistics sorting and small transport AGVs with a shorter overall path-planning time, higher stability, and limited computing power, and it has reference significance and practical value.

Traumatic Brain Injury Structure Detection Using Advanced Wavelet Transformation Fusion Algorithm with Proposed CNN-ViT

Detecting Traumatic Brain Injuries (TBI) through imaging remains challenging due to limited sensitivity in current methods. This study addresses the gap by proposing a novel approach integrating deep-learning algorithms and advanced image-fusion techniques to enhance detection accuracy. The method combines contextual and visual models to effectively assess injury status. Using a dataset of repeat mild TBI (mTBI) cases, we compared various image-fusion algorithms: PCA (89.5%), SWT (89.69%), DCT (89.08%), HIS (83.3%), and averaging (80.99%). Our proposed hybrid model achieved a significantly higher accuracy of 98.78%, demonstrating superior performance. Metrics including Dice coefficient (98%), sensitivity (97%), and specificity (98%) verified that the strategy is efficient in improving image quality and feature extraction. Additional validations with “entropy”, “average pixel intensity”, “standard deviation”, “correlation coefficient”, and “edge similarity measure” confirmed the robustness of the fused images. The hybrid CNN-ViT model, integrating curvelet transform features, was trained and validated on a comprehensive dataset of 24 types of brain injuries. The overall accuracy was 99.8%, with precision, recall, and F1-score of 99.8%. The “average PSNR” was 39.0 dB, “SSIM” was 0.99, and MI was 1.0. Cross-validation across five folds proved the model’s “dependability” and “generalizability”. In conclusion, this study introduces a promising method for TBI detection, leveraging advanced image-fusion and deep-learning techniques, significantly enhancing medical imaging and diagnostic capabilities for brain injuries.

Hyperspectral image classification with token fusion on GPU

Hyperspectral images capture material nuances with spectral data, vital for remote sensing. Transformer has become a mainstream approach for tackling the challenges posed by high-dimensional hyperspectral data with complex structures. However, a major challenge they face when processing hyperspectral images is the presence of a large number of redundant tokens, which leads to a significant increase in computational load, adding to the model’s computational burden and affecting inference speed. Therefore, we propose a token fusion algorithm tailored to the operational characteristics of the hyperspectral image and pure transformer network, aimed at enhancing the final accuracy and throughput of the model. The token fusion algorithm introduces a token merging step between the attention mechanism and the multi-layer perceptron module in each Transformer layer. Experiments on four hyperspectral image datasets demonstrate that our token fusion algorithm can significantly improve inference speed without any training, while only causing a slight decrease in the pure transformer network’s classification accuracy.

A modified Fuvar fusion algorithm based on adaptive end-member selection for hyperspectral remote sensing images

ABSTRACT A significant strategy for achieving a balance between spatial and spectral resolution is to combine hyperspectral remote sensing images with low spatial resolution and multispectral remote sensing images with high spatial resolution. For the issues of spectrum distortion and lack of the ability of gaining high frequency information when the Fuvar algorithm employs the Vertex Component Analysis (VCA) technique, this thesis presents a modified Fuvar (MFuvar) algorithm that utilizes the Maximum Distance Analysis (MDA) method instead of the VCA approach. The two subsets from the hyperspectral remote sensing image of GF-5 and the multispectral remote sensing image of Sentinel-2A, representing different land cover types were used as test data. The spectral fidelity and the ability of gaining high frequency information were assessed by using visual and statistical analysis. Fused images are compared with eight fusion methods, including SFIMHS, GLPHS, MAPSMM, CNMF, Hysure, SpaFusion, LTTR, and Fuvar, respectively. The results show that the MFuvar algorithm can keep the best balance between spectral fidelity and the ability of gaining high frequency information, and it is generally better than the compared eight algorithms. And it fulfils the automatic selection of end elements without manual intervention and increases the efficiency of algorithm operation.

The influence maximization algorithm for integrating attribute graph clustering and heterogeneous graph transformer

In social networks, maximizing influence is an important research direction. However, traditional influence maximization algorithms often overlook the attribute information of nodes and the heterogeneity of networks, leading to inefficiency and inaccuracy in the propagation process. To address this issue, this study first constructs a social network influence maximization propagation model, and then combines auto-encoder and graph convolutional autoencoder to extract social network user attributes. Finally, Transformer is used to learn the feature representation of social network nodes. Moreover, a heterogeneous graph neural network is introduced to combine with Transformer to further learn the feature representation of nodes, so as to design an influence maximization algorithm combining attributed graph clustering and heterogeneous graph Transformer. The results showed that the loss values of the fusion algorithm on different datasets were 0.619 and 0.17, respectively, proving its good fitting performance. The recall rate and F1 of this fusion algorithm were 92.5 % and 0.90, respectively, proving its high clustering accuracy. The influence maximization model based on fusion algorithm achieved active node coverage of 67 % and 48 % on two datasets, respectively, proving its good influence propagation effect. The above results demonstrate that the designed model can effectively spread influence. This helps to better understand and utilize the influence dissemination mechanisms in social networks, thereby promoting the development of social network research.

Construction and verification of pancreatic fistula risk prediction model after pancreaticoduodenectomy based on ensemble machine learning

Objective: To construct an ensemble machine learning model for predicting the occurrence of clinically relevant postoperative pancreatic fistula (CR-POPF) after pancreaticoduodenectomy and evaluate its application value. Methods: This is a research on predictive model. Clinical data of 421 patients undergoing pancreaticoduodenectomy in the Department of Pancreatic Surgery,Union Hospital, Tongji Medical College,Huazhong University of Science and Technology from June 2020 to May 2023 were retrospectively collected. There were 241 males (57.2%) and 180 females (42.8%) with an age of (59.7±11.0)years (range: 12 to 85 years).The research objects were divided into training set (315 cases) and test set (106 cases) by stratified random sampling in the ratio of 3∶1. Recursive feature elimination is used to screen features,nine machine learning algorithms are used to model,three groups of models with better fitting ability are selected,and the ensemble model was constructed by Stacking algorithm for model fusion. The model performance was evaluated by various indexes,and the interpretability of the optimal model was analyzed by Shapley Additive Explanations(SHAP) method. The patients in the test set were divided into different risk groups according to the prediction probability (P) of the alternative pancreatic fistula risk score system (a-FRS). The a-FRS score was validated and the predictive efficacy of the model was compared. Results: Among 421 patients,CR-POPF occurred in 84 cases (20.0%). In the test set,the Stacking ensemble model performs best,with the area under the curve (AUC) of the subject's work characteristic curve being 0.823,the accuracy being 0.83,the F1 score being 0.63,and the Brier score being 0.097. SHAP summary map showed that the top 9 factors affecting CR-POPF after pancreaticoduodenectomy were pancreatic duct diameter,CT value ratio,postoperative serum amylase,IL-6,body mass index,operative time,albumin difference before and after surgery,procalcitonin and IL-10. The effects of each feature on the occurrence of CR-POPF after pancreaticoduodenectomy showed a complex nonlinear relationship. The risk of CR-POPF increased when pancreatic duct diameter<3.5 mm,CT value ratio<0.95,postoperative serum amylase concentration>150 U/L,IL-6 level>280 ng/L,operative time>350 minutes,and albumin decreased by more than 10 g/L. The AUC of a-FRS in the test set was 0.668,and the prediction performance of a-FRS was lower than that of the Stacking ensemble machine learning model. Conclusion: The ensemble machine learning model constructed in this study can predict the occurrence of CR-POPF after pancreaticoduodenectomy,and has the potential to be a tool for personalized diagnosis and treatment after pancreaticoduodenectomy.

FusionOC: Research on optimal control method for infrared and visible light image fusion

Infrared and visible light image fusion can solve the limitations of single-type visual sensors and can boost the target detection performance. However, since the traditional fusion strategy lacks the controllability and feedback mechanism, the fusion model cannot precisely perceive the relationship between the requirements of the fusion task, the fused image quality, and the source image features. To this end, this paper establishes a fusion model based on the optimal controlled object and control mode called FusionOC. This method establishes two types of mathematical models of the controlled objects by verifying the factors and conflicts affecting the quality of the fused image. It combines the image fusion model with the quality evaluation function to determine the two control factors separately. At the same time, two proportional-integral-derivative (PID) control and regulation modes based on the backpropagation (BP) neural network are designed according to the control factor characteristics. The fusion system can adaptively select the regulation mode to regulate the control factor according to the user requirements or the task to make the fusion system perceive the connection between the fusion task and the result. Besides, the fusion model employs the feedback mechanism of the control system to perceive the feature difference between the fusion result and the source image, realize the guidance of the source image feature to the entire fusion process, and improve the fusion algorithm's generalization ability and intelligence level when handling different fusion tasks. Experimental results on multiple public datasets demonstrate the advantages of FusionOC over advanced methods. Meanwhile, the benefits of our fusion results in object detection tasks have been demonstrated.

Fusion Algorithm Based on Improved A* and DWA for USV Path Planning

Fusion Algorithm Based on Improved A* and DWA for USV Path Planning

Research on capacitance-coupled electrical impedance imaging fusion algorithm based on wavelet transform.

Capacitively Coupled Electrical Impedance Tomography (CCEIT), as a new electrical tomography method, effectively overcomes the drawbacks of traditional electrical resistance tomography that requires direct contact with liquids, so as to achieve non-contact measurement. In order to make full use of the complete electrical impedance information (real and imaginary) obtained by CCEIT and further improve the quality of image reconstruction, an image fusion algorithm based on wavelet transform was proposed. First, the Landweber algorithm is used to image the real part information and the imaginary part information, and the wavelet transform algorithm is used to decompose the two images at multiple scales, in which the low-frequency coefficient is fused by the improved logic filter algorithm based on the adaptive threshold, the high-frequency coefficient is selected based on the absolute value maximum method for fusion, and finally, the fused image is obtained by wavelet reconstruction. Numerical simulations and experiments verify the feasibility of the proposed method, and the results show that the relative error is reduced by up to 17% and the correlation coefficient is increased by up to 10%.

Infrared and visible image fusion based on hybrid multi-scale decomposition and adaptive contrast enhancement

Effectively fusing infrared and visible images enhances the visibility of infrared target information while capturing visual details. Balancing the brightness and contrast of the fusion image adequately has posed a significant challenge. Moreover, preserving detailed information in fusion images has been problematic. To address these issues, this paper proposes a fusion algorithm based on multi-scale decomposition and adaptive contrast enhancement. Initially, we present a hybrid multi-scale decomposition method aimed at extracting valuable information comprehensively from the source image. Subsequently, we advance an adaptive base layer optimization approach to regulate the brightness and contrast of the resultant fusion image. Lastly, we design a weight mapping rule grounded in saliency detection to integrate small-scale layers, thereby conserving the edge structure within the fusion outcome. Both qualitative and quantitative experimental results affirm the superiority of the proposed method over 11 state-of-the-art image fusion methods. Our method excels in preserving more texture and achieving higher contrast, which proves advantageous for monitoring tasks.

Research on the Fusion Positioning System of UWB/LiDAR Based on the Algorithm of SPF and KF

In order to overcome the disadvantages of single sensor localization in complex indoor environments, this paper proposes a design scheme for the ultra-wide band (UWB) and LiDAR co-location fusion system. To attain the ideal positioning effect of this system, a segmented point fusion approach based on Euclid's theorem is put forward to optimize the conventional extended Kalman filter and unscented Kalman filter algorithms. It can effectively reduce the positioning error of UWB in non-line-of-sight environments and under the multipath effect, and also assist LiDAR in correcting trajectory drift in sparsely textured scenes. Through both simulations and experiments, the feasibility of the proposed algorithm in the UWB/LiDAR positioning system is verified. The results reveal that the positioning error of the enhanced multi-sensor fusion algorithm is reduced by 22% compared with the original algorithm and 25.7% compared with the single sensor positioning method.

Recognizing Implicitly Toxic Content Based on Multiple Attention Mechanisms

ABSTRACTToxic content on social media has posed a significant threat to user experience and societal stability. In contrast to explicitly toxic content, implicitly toxic content, lacking overt toxic words, is challenging to identify through single‐text features. Therefore, this study proposes a multi‐feature fusion algorithm to recognize implicit toxic content. Firstly, we collect various features for each post, including text content, likes, comments, user information and other features. Subsequently, employing a multi‐head attention mechanism, we extract and fuse these features. Then, utilizing ensemble learning algorithms, we identify implicitly toxic content based on the fused features. Finally, we analyze the decision‐making process of the model using an interpretable algorithm and derive the most critical features and factors for identifying implicitly toxic content. The experimental results show that the proposed algorithm outperforms algorithms such as BERT, TextCNN, and XGBoost, demonstrating the advantage of multiple features over a single text feature in recognizing implicitly toxic content. In addition, this study provides insights into the decision‐making process of the model and provides more effective toxic content management strategies for social media platforms.

Leukocyte detection based on convolutional neural network fusion with color Fourier ptychographic microscopy

Abstract In the field of optical microscopic imaging, color Fourier Ptychographic Microscopy (FPM) technology has attracted much attention due to its advantages of large field of view, high resolution, and quantitative phase imaging. In this paper, a color FPM fusion algorithm based on deep learning is proposed in combination with Convolutional Neural Networks (CNN) and applied to leukocyte detection. Firstly, this paper introduces a fusion model of a convolutional neural network based on the traditional color FPM imaging method and fuses low-resolution color images and high-resolution grayscale images through a multilayer convolutional network. This method improves the quality of reconstructed images while reducing the reconstruction time. Secondly, this paper constructs a leukocyte detection dataset by using an improved color FPM reconstruction algorithm and builds a leukocyte detection system based on the YOLOv7 architecture. This paper shows that combining convolutional neural networks with color FPM technology can provide higher-quality reconstructed images in medical imaging and cell detection, which provides strong technical support for digital pathology and medical diagnosis.

Multimodal biometric integration: Trends and insights from the past quinquennial

Recent years have seen a notable breakthrough in multimodal biometric systems, which combine numerous biometric modalities for increased security and accuracy. This paper offers a thorough analysis of recent advances in the field, addressing the many multimodal biometric strategies that have been put out by different researchers. Numerous modalities are covered by the evaluation, such as palm print, speech, iris, facial features, palm print, fingerprint, body form, gait, and more. To obtain amazing results in terms of accuracy and security, researchers have used cutting-edge methods like convolutional neural networks (CNNs), support vector machines (SVM), recurrent neural networks (RNNs), deep learning, and optimization algorithms. Many fusion strategies have been investigated to efficiently merge data from many modalities, such as feature-level fusion, decision-level fusion, and score-level fusion. Developments in template protection systems have also addressed security issues related to transmission and storage of biometric data. Even though a lot of the suggested systems have shown great accuracy rates, there are still issues with hardware restrictions, dataset biases, privacy problems, and computational costs. Prospective study avenues encompass investigating more extensive and heterogeneous datasets, crafting resilient fusion algorithms, and amalgamating cutting-edge technologies like deep learning-based biometric cryptosystems and federated learning. All things considered, the literature study demonstrates the enormous potential of multimodal biometric systems in offering safe and dependable authentication solutions for a wide range of applications, from on-line student authentication and proctoring systems to customized healthcare networks.

Sowing Depth Monitoring System for High-Speed Precision Planters Based on Multi-Sensor Data Fusion

High-speed precision planters are subject to high-speed (12~16 km/h) operation due to terrain undulation caused by mechanical vibration and sensor measurement errors caused by the sowing depth monitoring system’s accuracy reduction problems. Thus, this study investigates multi-sensor data fusion technology based on the sowing depth monitoring systems of high-speed precision planters. Firstly, a sowing depth monitoring model comprising laser, ultrasonic, and angle sensors as the multi-sensor monitoring unit is established. Secondly, these three single sensors are filtered using the Kalman filter. Finally, a multi-sensor data fusion algorithm for optimising four key parameters in the extended Kalman filter (EKF) using an improved sparrow search algorithm (ISSA) is proposed. Subsequently, the filtered data from the three single sensors are integrated to address the issues of mechanical vibration interference and sensor measurement errors. In order to ascertain the superiority of the ISSA-EKF, the ISSA-EKF and SSA-EKF are simulated, and their values are compared with the original monitoring value of the sensor and the filtered sowing depth value. The simulation test demonstrates that the ISSA-EKF-based sowing depth monitoring algorithm for high-speed precision planters, with a mean absolute error (MAE) of 0.083 cm, root mean square error (RMSE) of 0.103 cm, and correlation coefficient (R) of 0.979 achieves high-precision monitoring. This is evidenced by a significant improvement in accuracy when compared with the original monitoring value of the sensor, the filtered value, and the SSA-EKF. The results of a field test demonstrate that the ISSA-EKF-based sowing depth monitoring system for high-speed precision planters enhances the precision and reliability of the monitoring system when compared with the three single-sensor monitoring values. The average MAE and RMSE are reduced by 0.071 cm and 0.075 cm, respectively, while the average R is improved by 0.036. This study offers a theoretical foundation for the advancement of sowing depth monitoring systems for high-speed precision planters.

Robust fusion filter for networked uncertain descriptor systems with colored noise and cyber-attacks

The robust fusion filtering problem of multi-sensor networked uncertain descriptor systems (NUDSs) with colored noise, uncertain noise variances and cyber-attacks is investigated. During data transmission in unreliable communication networks, the data can be maliciously attacked by attackers. In other words, the local filters (LFs) may receive false data or may not receive data because of the cyber-attacks. By adopting the singular value decomposition (SVD) method, the original NUDSs can be converted into two reduced-order subsystems with uncertain correlated fictitious white noises, and the cyber-attacks are transformed into the fictitious noises. Cross-covariance matrices between local filtering errors are derived. The robust LFs are obtained according to the minimax robust estimation principle. Under the linear unbiased minimum variance criterion, three weighted fusion algorithms are applied to fuse the LFs. For all allowable uncertainties of noise variances and cyber-attacks, the minimal upper bounds of covariance matrices of the local and distributed fusion filters are guaranteed. The proof of their robustness is established through the minimax estimation principle and Lyapunov equation method. Finally, the correctness and effectiveness of the proposed algorithms are verified by a circuit system example.

Multi-Focus Image Fusion Using Energy Valley Optimization Algorithm

When a natural scene is photographed using imaging sensors commonly used today, part of the image is obtained sharply while the other part is obtained blurry. This problem is called limited depth of field. This problem can be solved by fusing the sharper parts of multi-focus images of the same scene. These methods are called multi-focus image fusion methods. This study proposes a block-based multi-focus image fusion method using the Energy Valley Optimization Algorithm (EVOA), which has been introduced in recent years. In the proposed method, the source images are first divided into uniform blocks, and then the sharper blocks are determined using the criterion function. By fusing these blocks, a fused image is obtained. EVOA is used to optimize the block size. The function that maximizes the quality of the fused image is used as the fitness function of the EVOA. The proposed method has been applied to commonly used image sets. The obtained experimental results are compared with the well-known Genetic Algorithm (GA), Differential Evolution Algorithm (DE), and Artificial Bee Colony Optimization Algorithm (ABC). The experimental results show that EVOA can compete with the other block-based multi-focus image fusion algorithms.

Neural-aware Decoupling Fusion based Personalized Federated Learning for Intelligent Sensing

Personalized federated learning (PFL) framework that targets individual models for optimization provides better privacy and flexibility for clients. However, in challenging intelligent sensing applications, the heterogeneous client’s data distributions make the aggregation of local models in the server unstable or even hard to converge. To deal with the performance degradation caused by the above problem, existing PFL methods focus more on how to fine-tune the global model but ignore the impact of the global model fusion algorithm on the results. In this paper, we propose a new explainable neural-aware decoupling fusion based PFL framework, p-FedADF , to address the above challenges. It contains two carefully designed modules. The local decoupling module deployed on the client, utilizes the architecture disentangle technique to decouple the feature extractors in the client’s local model into sub-network according to data categories. It obtains the inference process of feature extraction for different categories of data by training. The global aggregation module, deployed on the server, aligns the sub-networks positions for multiple clients and implements a fine-grained generic feature extractor aggregation. In addition, we provide a mask encoding scheme to reduce the communication overhead of transmitting the sub-network sets between the server and clients. Our p-FedADF obtains 1.6%, 0.2%, 2.3%, and 4.5% improvement on a real-world dataset and three benchmark datasets, compared to state-of-the-art (SOTA) methods.

Improved Maximum Correntropy Criterion Kalman Filter with Adaptive Behaviors for INS/UWB Fusion Positioning Algorithm

Accurate indoor robot navigation cannot be achieved without reliable indoor positioning techniques. Ultra-wideband (UWB) is among the most dependable techniques currently available. However, the complexity of indoor environments results in signal transmissions that are susceptible to interference from obstacles, which in turn reduces positioning accuracy. Inertial Navigation System (INS) is an autonomous navigation system that is free from interference in indoor environments and unaffected by non-line-of-sight (NLOS) conditions. This paper proposes a new joint INS and UWB positioning method utilizing the Maximum Correntropy Criterion Kalman Filter (MCCKF). This approach effectively cope with the interference of measurement outliers and extend the design of the adaptive mechanisms to enhance the performance of the localization system. For UWB positioning of tag nodes, an improved Particle Swarm Optimization combined with kmeans (PSO-kmeans) method is used to reduce the impact of NLOS errors on positioning. Finally, the INS is calibrated by AMCCKF fused positioning results. The results of simulations and experiments demonstrate that the proposed AMCCKF fusion algorithm effectively suppresses the impact of anomalous measurements, enhances positioning accuracy and robustness, thereby improving its practicality in real-world environments.