Aiguo Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Aiguo Li has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Jiahao Fu
School of Computer Science and Technology, Xi’an University of Science and Technology, Xi’an 710054, China
2 shared publications - Jie Yang
College of Computer Science and Technology, Xi'an University of Science and Technology, China
2 shared publications - Ruifang Gao
School of Computer Science and Technology, Xi’an University of Science and Technology, Xi’an 710054, China
2 shared publications Chen Yang
College of Computer Science and Technology, Xi'an University of Science and Technology, Xi'an, China
1 shared publication- Defeng Wu
Jimei University
1 shared publication - Fan Yang
College of Computer Science and Technology, Xi'an University of Science and Technology, Xi'an, Shaanxi, P.R. China
1 shared publication
Latest Publications
- Open Access
- 24 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 37457-37466
A 3D Most Vulnerable Path Search Method for Physical Protection Systems Based on the EASI Model
Physical Protection System (PPS) is a crucial system for critical infrastructure to protect its vital information, and it is essential to analyze the vulnerability of PPS. Established methods for PPS vulnerability analysis mainly focus on one- and two-dimensional environments. However, the air and underwater threats are increasing, and more effective methods are needed to deal with these new threats than the existing methods. The security of the physical protection system in three-dimensional space needs to be analyzed. The most vulnerable path search is a critical link in the security assessment of the physical protection system, so this paper proposes a three-dimensional physical protection system based on the Estimate of Adversary Sequence Interruption(EASI) model of the most vulnerable path search method. A three-dimensional space model is established, and the algorithm’s search points are reduced based on existing methods. The search efficiency is improved by restricting the intermediate points and the angle between the start and target points and improving the heuristic function. The improved algorithms are simulated and compared, and the results show that the paths generated by the improved algorithms are more reasonable, and the algorithms consume less time and have shorter path lengths.
- 24 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 37457-37466
Open AccessA 3D Most Vulnerable Path Search Method for Physical Protection Systems Based on the EASI Model
Physical Protection System (PPS) is a crucial system for critical infrastructure to protect its vital information, and it is essential to analyze the vulnerability of PPS. Established methods for PPS vulnerability analysis mainly focus on one- and two-dimensional environments. However, the air and underwater threats are increasing, and more effective methods are needed to deal with these new threats than the existing methods. The security of the physical protection system in three-dimensional space needs to be analyzed. The most vulnerable path search is a critical link in the security assessment of the physical protection system, so this paper proposes a three-dimensional physical protection system based on the Estimate of Adversary Sequence Interruption(EASI) model of the most vulnerable path search method. A three-dimensional space model is established, and the algorithm’s search points are reduced based on existing methods. The search efficiency is improved by restricting the intermediate points and the angle between the start and target points and improving the heuristic function. The improved algorithms are simulated and compared, and the results show that the paths generated by the improved algorithms are more reasonable, and the algorithms consume less time and have shorter path lengths.
- Open Access
- 6 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 117-122
Improved safety helmet target detection algorithm of YOLOv8n
- 6 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 117-122
Open AccessImproved safety helmet target detection algorithm of YOLOv8n
- Open Access
- 6 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 123-130
Ultrasonic Metal Recognition Based on a Time-Frequency Feature Multi-Scale Fusion Network
When identifying metals with the same appearance and material non - destructively, the existing ultrasonic metal identification methods often overlook unknown - class identification, have poorly constructed data sets and limited practicality. This paper proposes an ultrasonic metal identification method based on the multi - scale fusion network of time - frequency features. It uses the multi - head multi - scale residual convolution module to extract and fuse time - frequency features, and combines the attention mechanism to process the fused features, which overcomes the over - fitting problem of a single time - domain feature and improves the generalization ability. An independent data set including training, validation and test sets is constructed. The training and validation sets each contain ultrasonic data of 12 test blocks, and the test set contains 13 classes (the 13th class is 5 test blocks of unknown categories). The test - set data are not used in model training to simulate actual scenarios. Experiments show that this method can efficiently identify metals in the self - built data set, and its effectiveness in classifying long sequences is verified on the UCR data set.
- 6 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 123-130
Open AccessUltrasonic Metal Recognition Based on a Time-Frequency Feature Multi-Scale Fusion Network
When identifying metals with the same appearance and material non - destructively, the existing ultrasonic metal identification methods often overlook unknown - class identification, have poorly constructed data sets and limited practicality. This paper proposes an ultrasonic metal identification method based on the multi - scale fusion network of time - frequency features. It uses the multi - head multi - scale residual convolution module to extract and fuse time - frequency features, and combines the attention mechanism to process the fused features, which overcomes the over - fitting problem of a single time - domain feature and improves the generalization ability. An independent data set including training, validation and test sets is constructed. The training and validation sets each contain ultrasonic data of 12 test blocks, and the test set contains 13 classes (the 13th class is 5 test blocks of unknown categories). The test - set data are not used in model training to simulate actual scenarios. Experiments show that this method can efficiently identify metals in the self - built data set, and its effectiveness in classifying long sequences is verified on the UCR data set.
- Open Access
- 6 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 112-116
Multi-Scene Fire Smoke Detection Algorithm Based on YOLO v8
- 6 December 2024
- Published by Association for Computing Machinery (ACM)
- p. 112-116
Open AccessMulti-Scene Fire Smoke Detection Algorithm Based on YOLO v8
- Open Access
- 6 December 2023
- Published by Association for Computing Machinery (ACM)
- p. 306-311
Multi-sensor data fusion method based on adaptive Kalman filtering
Since the data collected by a single sensor often cannot meet the needs and is far from forming a comprehensive and complete perceptual description of complex systems, multi-sensor data fusion technology is required to comprehensively process the data and obtain more accurate results. The application of Kalman filter to multi-sensor data fusion has become a hot topic in recent years, but when the noise matrix design of Kalman filter is unreasonable, filtering divergence is prone to occur, so an adaptive Kalman filter algorithm is proposed, which uses correction coefficients to correct the filtering process by obtaining the dynamic change of the residual in real time. The anti-collision system of ship unloader is used as the application background to verify the effectiveness of the proposed algorithm. Experimental results show that the proposed algorithm improves the estimation accuracy and reduces noise interference in the actual environment.
- 6 December 2023
- Published by Association for Computing Machinery (ACM)
- p. 306-311
Open AccessMulti-sensor data fusion method based on adaptive Kalman filtering
Since the data collected by a single sensor often cannot meet the needs and is far from forming a comprehensive and complete perceptual description of complex systems, multi-sensor data fusion technology is required to comprehensively process the data and obtain more accurate results. The application of Kalman filter to multi-sensor data fusion has become a hot topic in recent years, but when the noise matrix design of Kalman filter is unreasonable, filtering divergence is prone to occur, so an adaptive Kalman filter algorithm is proposed, which uses correction coefficients to correct the filtering process by obtaining the dynamic change of the residual in real time. The anti-collision system of ship unloader is used as the application background to verify the effectiveness of the proposed algorithm. Experimental results show that the proposed algorithm improves the estimation accuracy and reduces noise interference in the actual environment.
- Open Access
- 1 January 2023
- Published by American Institute of Mathematical Sciences (AIMS) in Mathematical Biosciences and Engineering
- Vol. 20 (9) , 16744-16761
AGMG-Net: Leveraging multiscale and fine-grained features for improved cargo recognition
Security systems place great emphasis on the safety of stored cargo, as any loss or tampering can result in significant economic damage. The cargo identification module within the security system faces the challenge of achieving a 99.99% recognition accuracy. However, current identification methods are limited in accuracy due to the lack of cargo data, insufficient utilization of image features and minimal differences between actual cargo classes. First, we collected and created a cargo identification dataset named "Cargo" using industrial cameras. Subsequently, an Attention-guided Multi-granularity feature fusion model (AGMG-Net) was proposed for cargo identification. This model extracts both coarse-grained and fine-grained features of the cargo using two branch networks and fuses them to fully utilize the information contained in these features. Furthermore, the Attention-guided Multi-stage Attention Accumulation (AMAA) module is introduced for target localization, and the Multi-region Optimal Selection method Based on Confidence (MOSBC) module is used for target cropping. The features from the two branches are fused using a fusion branch in a Concat manner for multi-granularity feature fusion. The experimental results show that the proposed model achieves an average recognition rate of 99.58, 92.73 and 88.57% on the self-built dataset Cargo, and the publicly available datasets Flower and Butterfly20, respectively. This is better than the state-of-the-art model. Therefore, this research method accurately identifies cargo categories and provides valuable assistance to security systems.
- 1 January 2023
- Published by American Institute of Mathematical Sciences (AIMS) in Mathematical Biosciences and Engineering
- Vol. 20 (9) , 16744-16761
Open AccessAGMG-Net: Leveraging multiscale and fine-grained features for improved cargo recognition
Security systems place great emphasis on the safety of stored cargo, as any loss or tampering can result in significant economic damage. The cargo identification module within the security system faces the challenge of achieving a 99.99% recognition accuracy. However, current identification methods are limited in accuracy due to the lack of cargo data, insufficient utilization of image features and minimal differences between actual cargo classes. First, we collected and created a cargo identification dataset named "Cargo" using industrial cameras. Subsequently, an Attention-guided Multi-granularity feature fusion model (AGMG-Net) was proposed for cargo identification. This model extracts both coarse-grained and fine-grained features of the cargo using two branch networks and fuses them to fully utilize the information contained in these features. Furthermore, the Attention-guided Multi-stage Attention Accumulation (AMAA) module is introduced for target localization, and the Multi-region Optimal Selection method Based on Confidence (MOSBC) module is used for target cropping. The features from the two branches are fused using a fusion branch in a Concat manner for multi-granularity feature fusion. The experimental results show that the proposed model achieves an average recognition rate of 99.58, 92.73 and 88.57% on the self-built dataset Cargo, and the publicly available datasets Flower and Butterfly20, respectively. This is better than the state-of-the-art model. Therefore, this research method accurately identifies cargo categories and provides valuable assistance to security systems.
- Open Access
- 1 December 2022
- Published by Association for Computing Machinery (ACM)
- p. 85-89
Research on vulnerability assessment of physical protection system based on combined weighting-two-dimensional cloud model
- 1 December 2022
- Published by Association for Computing Machinery (ACM)
- p. 85-89
Open AccessResearch on vulnerability assessment of physical protection system based on combined weighting-two-dimensional cloud model
- Open Access
- 1 December 2022
- Published by Association for Computing Machinery (ACM)
- p. 110-116
Multi-objective optimization algorithm for indoor positioning sensor deployment based on wireless network
Aiming at the influence of indoor buildings and sensor positions on signal strength, the optimization problem of wireless positioning sensor deployment is studied, and an optimization algorithm for positioning sensor deployment based on building signal penetration loss is proposed. First, define the minimum signal reception distance and rasterize the positioning two-dimensional space. Secondly, according to the characteristics of ranging and positioning algorithm, the concepts of positioning sensor deployment, effective signal coverage and building penetration loss are given. Then, a multi-objective optimization model of positioning sensor deployment based on building signal penetration loss is proposed with the optimization objectives of the number of sensors, effective coverage rate and overall coverage rate. Finally, the solution algorithm is designed based on NSGA-III, and compared with the uniform deployment, the deployment without considering the obstacles and the deployment of the Boolean perception model considering the obstacles. The simulation results show that the overall coverage rate of the algorithm under the same number is increased to 98.3%, and the effective coverage rate is increased to 92.5%; the number of positioning base stations under the same coverage rate is reduced, which reduces the positioning cost.
- 1 December 2022
- Published by Association for Computing Machinery (ACM)
- p. 110-116
Open AccessMulti-objective optimization algorithm for indoor positioning sensor deployment based on wireless network
Aiming at the influence of indoor buildings and sensor positions on signal strength, the optimization problem of wireless positioning sensor deployment is studied, and an optimization algorithm for positioning sensor deployment based on building signal penetration loss is proposed. First, define the minimum signal reception distance and rasterize the positioning two-dimensional space. Secondly, according to the characteristics of ranging and positioning algorithm, the concepts of positioning sensor deployment, effective signal coverage and building penetration loss are given. Then, a multi-objective optimization model of positioning sensor deployment based on building signal penetration loss is proposed with the optimization objectives of the number of sensors, effective coverage rate and overall coverage rate. Finally, the solution algorithm is designed based on NSGA-III, and compared with the uniform deployment, the deployment without considering the obstacles and the deployment of the Boolean perception model considering the obstacles. The simulation results show that the overall coverage rate of the algorithm under the same number is increased to 98.3%, and the effective coverage rate is increased to 92.5%; the number of positioning base stations under the same coverage rate is reduced, which reduces the positioning cost.
- Open Access
- 31 July 2022
- Published by Hindawi Limited in Scientific Programming
- Vol. 2022, 1-11
Semantic Trajectory Frequent Pattern Mining Method with Fuzzy Stay Time Constraint
In the security system, transforming a large number of collected target trajectories into semantic trajectories with a less volume and high quality and mining their frequent patterns are helpful to analyze the target behavior patterns, identify hazard sources, and enhance the internal prevention, and control of the security system. Aiming at the limitation of semantic trace frequent pattern mining method defined by precise stay time in practical application scenarios, a fuzzy semantic trace frequent pattern mining method is proposed. Firstly, the membership function of fuzzy stay time is defined, so the stay time of the target at the stay point is fuzzified, and the fuzzy semantic trajectory is obtained. Then, a fuzzy semantic trajectory frequent pattern mining algorithm FST-FPM (fuzzy semantic trajectory frequent pattern mining) is proposed. The FST-FPM algorithm is experimentally verified on the Geolife public dataset and the self-collected RFID positioning dataset. The experimental results show that FST-FPM algorithm can mine frequent patterns of fuzzy semantic trajectories on Geolife dataset and RFID positioning dataset, and the running time is reduced by more than 10 compared with classical PrefixSpan algorithm, PrefixSpan-x algorithm, and LFFT2 algorithm.
- 31 July 2022
- Published by Hindawi Limited in Scientific Programming
- Vol. 2022, 1-11
Open AccessSemantic Trajectory Frequent Pattern Mining Method with Fuzzy Stay Time Constraint
In the security system, transforming a large number of collected target trajectories into semantic trajectories with a less volume and high quality and mining their frequent patterns are helpful to analyze the target behavior patterns, identify hazard sources, and enhance the internal prevention, and control of the security system. Aiming at the limitation of semantic trace frequent pattern mining method defined by precise stay time in practical application scenarios, a fuzzy semantic trace frequent pattern mining method is proposed. Firstly, the membership function of fuzzy stay time is defined, so the stay time of the target at the stay point is fuzzified, and the fuzzy semantic trajectory is obtained. Then, a fuzzy semantic trajectory frequent pattern mining algorithm FST-FPM (fuzzy semantic trajectory frequent pattern mining) is proposed. The FST-FPM algorithm is experimentally verified on the Geolife public dataset and the self-collected RFID positioning dataset. The experimental results show that FST-FPM algorithm can mine frequent patterns of fuzzy semantic trajectories on Geolife dataset and RFID positioning dataset, and the running time is reduced by more than 10 compared with classical PrefixSpan algorithm, PrefixSpan-x algorithm, and LFFT2 algorithm.
- 1 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 32-35
Research on the target cargo identification method based on CNN
The automatic inspection robot can effectively pre-vent cargo from being stolen, replaced and opened without authorization. However, it will be disturbed by external factors such as occlusion, illumination and angle when it takes photos to identify cargo, which will affect the recognition effect. In order to improve the recognition effect of the automatic inspection robot on the target goods, this paper first detects the object in the data preprocessing stage, extracts the goods con-tour, and then carries on the image recognition, which im-proves the recognition rate. Then we compare the loss rate and accuracy rate of four network models: AlexNet, GoogleNet, ResNet and MobileNet. The experimental results show that ResN et algorithm has the highest image recognition accuracy, is more friendly to the recognition of light and occlusion, and is more in line with the actual operation needs of automatic inspection robot. It has reference value for the research of target cargo recognition.
- 1 December 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 32-35
Research on the target cargo identification method based on CNN
The automatic inspection robot can effectively pre-vent cargo from being stolen, replaced and opened without authorization. However, it will be disturbed by external factors such as occlusion, illumination and angle when it takes photos to identify cargo, which will affect the recognition effect. In order to improve the recognition effect of the automatic inspection robot on the target goods, this paper first detects the object in the data preprocessing stage, extracts the goods con-tour, and then carries on the image recognition, which im-proves the recognition rate. Then we compare the loss rate and accuracy rate of four network models: AlexNet, GoogleNet, ResNet and MobileNet. The experimental results show that ResN et algorithm has the highest image recognition accuracy, is more friendly to the recognition of light and occlusion, and is more in line with the actual operation needs of automatic inspection robot. It has reference value for the research of target cargo recognition.