Loading...

Table of Content

    29 November 2024, Volume 0 Issue 11
    Abstract Tree Model for Gridded Cube Metadata
    LI Deyou1, 2, YU Jinsongdi1, 2, WEI Dandan1, 2, LUO Yuan1, 2, TONG Ruiju3
    2024, 0(11):  1-6.  doi:10.3969/j.issn.1006-2475.2024.11.001
    Asbtract ( 50 )   PDF (3488KB) ( 57 )  
    References | Related Articles | Metrics
    This paper aimed to investigate an abstract tree model for gridded cube metadata, designed to facilitate the management of heterogeneous multisource earth observation metadata, thereby reducing the need for parsing, adaptation, and storage structure design for heterogeneous metadata. Firstly, based on the analysis of the fundamental characteristics of earth observation metadata, this study abstracted the general components of metadata and defined a unified abstract tree model along with its operational methods. Subsequently, the storage and operational mapping of abstract tree models were implemented based on a relational database system. This endeavor was complemented by the design of a gridded cube metadata model aligned with prevalent international standards for geographic information interoperation. The overarching objective was to achieve unified integration and storage of heterogeneous metadata. Finally, a prototype system for managing gridded cube metadata was constructed. The management of North Atlantic environmental variables data was presented as an application case to demonstrate the operations of gridded cube metadata based on the abstract tree model. The application indicates that the proposed metadata model is user-friendly, highly extensible, and effectively supports the integration and storage of earth observation metadata.
    Beijing Opera Binary Classification Based on RF-LCE-BiLSTM-Attention-AMSSA Model 
    GONG Yicheng1, 2, LIU Qing1, 2
    2024, 0(11):  7-12.  doi:10.3969/j.issn.1006-2475.2024.11.002
    Asbtract ( 42 )   PDF (1380KB) ( 43 )  
    References | Related Articles | Metrics
     In order to improve the classification accuracy of Beijing Opera in the era of big data and promote the dissemination of national essence, this article uses the deviation penalty cross entropy loss function on the basis of RF-BiLSTM-Attention to prevent overfitting and integrates the Adaptive Multi-Swarm Sparrow Search Algorithm (AMSSA) to propose the RF-LCE-BiLSTM-Attention-AMSSA model for binary classification of Beijing Opera and other music. The model first converts audio files into feature vectors, and then combines L2 regularization loss and Cross Entropy loss (LCE) as the deviation penalty cross entropy loss function of the model, which is trained through neural networks for classification. After that, the AMSSA is adopted to optimize the hyperparameters, and the optimal hyperparameters are applied for the binary classification of Beijing Opera. A Beijing Opera binary classification experiment was conducted on 1500 pieces of music, which come from the popular music platform and GTZAN dataset, to compare the classification accuracy of RF-LCE-BiLSTM-Attention-AMSSA model with 11 models such as RNN, LSTM, and BiLSTM, and to compare the impact of LCE loss function and AMSSA on the model. The results show that the classification accuracy of RF-LCE-BiLSTM-Attention-AMSSA model is 89.95%, which is 0.95 percentage points higher than RF-BiLSTM-Attention, and 0.28 percentage points higher than RF-LCE-BiLSTM-Attention. 
    Taxi Passenger Flow Prediction Based on Heterogeneous Spatiotemporal Graph#br# Convolutional Networks 
    LI Taoying, LI Meng, WU Mengqiao
    2024, 0(11):  13-18.  doi:10.3969/j.issn.1006-2475.2024.11.003
    Asbtract ( 53 )   PDF (3015KB) ( 46 )  
    References | Related Articles | Metrics
     Accurately predicting regional taxi passenger flow plays an important role in taxi dispatch and passenger transportation. The exploration of spatiotemporal correlations in passenger flow is a critical factor in enhancing prediction accuracy. In light of the limited investigation into the spatiotemporal characteristics of regional passenger flow, particularly the inadequate exploration of passenger flow similarities between non-adjacent regions and the underexplored spatial relationships among regions, a Heterogeneous Spatio-Temporal Graph Convolutional Network (HSTGCN) is proposed to predict the passenger flow across multiple target regions. To capture the spatiotemporal characteristics of passenger flow data, we construct a heterogeneous graph utilizing regional physical adjacency graphs, regional similarity graphs, and origin-destination (OD) correlation graphs. Furthermore, based on these adjacency matrices, we build a dynamic graph reflecting the spatiotemporal dynamics of regions. We employ heterogeneous spatiotemporal graph convolutional networks to extract the spatiotemporal features of the data. Experimental results on publicly available datasets demonstrate that the model’s prediction outcomes outperform comparative models in terms of mean absolute error, root mean square error, accuracy, and R2, showcasing superior prediction accuracy. 
    Information Forwarding Strategy of Internet of Vehicles in Named Data Network 
    ZHANG Tai1, YAN Zihao2, DUAN Jie2, ZHANG Zhihong2
    2024, 0(11):  19-27.  doi:10.3969/j.issn.1006-2475.2024.11.004
    Asbtract ( 46 )   PDF (1564KB) ( 45 )  
    References | Related Articles | Metrics
    Aiming at the problem of dynamic change of network topology due to high mobility of nodes and the potential broadcast storm problem of traditional networks in Internet of Vehicles, a network topology prediction-based information forwarding strategy for Internet of Vehicles is proposed. Firstly, according to the performance of nodes, vehicle operation state attributes, movement trajectory, link attributes, etc. in the Internet of Vehicles, a dynamic network topology evolution algorithm is used for topology prediction to construct the current network topology; and then a multi-source information forwarding table is constructed using the shortest path algorithm, and information forwarding is performed according to the information forwarding table. In the case of forwarding table failure, the best neighboring nodes are selected to forward data based on node and link attributes, and at the same time, the topology prediction-based content forwarding is further improved. Simulation results show that compared with other forwarding strategies, the proposed forwarding strategy can effectively reduce the content acquisition delay and improve the request hit rate.
    Inventory Forecasting Method Based on Improved Elman Neural Network
    YUAN Qingle, MU Li
    2024, 0(11):  28-33.  doi:10.3969/j.issn.1006-2475.2024.11.005
    Asbtract ( 42 )   PDF (1515KB) ( 58 )  
    References | Related Articles | Metrics
    Because the procurement management of steel enterprises lacks reasonable planning in cyclical procurement amounts, the production demand forecasts for supply enterprises are inaccurate. To address these issues, a model based on improved cuckoo search algorithm to optimize Elman neural network (BASCS-Elman) is proposed. Taking material iron ore of Desheng Company of Baosteel as the research object, this model is used to predict demand to achieve accurate prediction, reduce resource waste, and improve enterprise profits. In this paper, the initial CS population is optimized by Logistic chaotic mapping to maintain the diversity of the population and improve the uniformity of the algorithm’s search. The traversal global search capability is increased by updating cuckoo locations through adaptive Levy flight. The multi-stage dynamic disturbance strategy helps global optimization. The local optimization speed is accelerated by the cow whiskers beetle antennae search algorithm. Finally the simulation experiment results show that, the average absolute error of the proposed model is 1.5042, the average absolute percentage error is 0.33423%, and the fastest stable time is 1.18 s, which is better than other prediction models.
    Q-learning-based Algorithm for Orchestrating Security Service Function Chain
    LIU Xing1, 2, GUO Liang1, 2, WANG Zhengqi1, 2, WEI Xiaogang1, 2, XU Xuefei1, 2, LIU Jing3
    2024, 0(11):  34-40.  doi:10.3969/j.issn.1006-2475.2024.11.006
    Asbtract ( 45 )   PDF (1257KB) ( 38 )  
    References | Related Articles | Metrics
    With the development of technology, Internet is becoming an indispensable part of human life and network security is becoming particularly important. To ensure network security, the orchestration of dynamic security service function chains is an important research direction. However, current research on network resource mapping and orchestration algorithms for dynamic security service function chains mainly focuses on a specific type of network resource, with the main goal of optimizing a certain network resource and reducing network service latency. They overlook the balance of overall resource allocation in the network. We construct a physical network model and a security service function chain model. Considering both physical network node computing resources and link bandwidth resources while meeting user needs, the goal is to achieve the best-balanced allocation of network resources. Based on the reinforcement Q-learning algorithm, a new link arrangement reward method is proposed, and a greedy strategy is introduced to avoid falling into local optima. A typical physical network model and different numbers of security service function chains that needs to be arranged are selected and the optimal arrangement path of the security service function chain is obtained through multiple iterations. The simulation results show that the optimal arrangement of the proposed security service function chain reduces the arrangement response time by 38.5% and improves the resource allocation balance by 2.1% compared to the simulated annealing algorithm. Compared with a genetic algorithm, it reduces the response time of orchestration by 96.5% and improves the balance of resource allocation by 2.9%.
    Zero-trust Dynamic Evaluation Method for IoT Terminals
    DONG Chongchong, ZHAO Cong, WU You, ZHANG Lei, ZHANG Jiawen, LI Zhihao
    2024, 0(11):  41-45.  doi:10.3969/j.issn.1006-2475.2024.11.007
    Asbtract ( 56 )   PDF (1553KB) ( 34 )  
    References | Related Articles | Metrics
     The zero trust network security architecture is committed to ensure the access security of Internet of Things (IoT) terminal devices. However, the heterogeneous nature of devices and the real-time nature of the data received by the network lead to the increase of network attacks and cannot be effectively defended. Therefore, we propose a method that can effectively and actively determine safety. This paper introduces the idea of rate of change in mathematics into trust analysis, and forms three attribute sets based on trust interval and rate of change: discrete interval, change range, and change frequency. By calculating the above attributes of the entity’s trust value, the entity’s trust situation is obtained, and an overall assessment of the terminal entity’s trust situation is made from the three levels of completeness, accuracy and objectivity. Under the premise of reducing encryption and other means, the above method can evaluate the trust state of the IoT terminal from the perspective of the data, and this evaluation method can provide a basis for the judgment of the IoT terminal more objectively and accurately. 
    An LLM-based Method for Automatic Construction of Equipment Failure Knowledge Graphs
    ZHANG Kun1, ZHANG Yongwei1, WU Yongcheng1, ZHANG Xiaowen2, ZHAI Shichen2
    2024, 0(11):  46-53.  doi:10.3969/j.issn.1006-2475.2024.11.008
    Asbtract ( 111 )   PDF (5470KB) ( 96 )  
    References | Related Articles | Metrics
    Fault operation and maintenance is an important research topic in the field of industrial production. The research of fault prediction, fault diagnosis, question-answering systems based on the fault knowledge graph have been greatly developed and applied. However, a high-quality fault operation and maintenance knowledge graph is the foundation for these methods. Considering that traditional knowledge graph construction techniques require data preprocessing, entity recognition, relationship extraction and entity alignment of raw data, to improve the efficiency of knowledge graphs, this paper focuses on using large language models for unsupervised knowledge extraction from fault operation and maintenance data to achieve automatic construction of large-scale fault operation and maintenance knowledge graphs. This method mainly includes two parts: 1) Two zero-shot Prompts oriented towards the construction of fault operation and maintenance knowledge graphs are proposed. These Prompts can guide large language models to generate conceptual layers and extract elemental knowledge for the fault operation and maintenance knowledge graph represented and output in RDF syntax; 2) A method based on large language models for constructing knowledge graphs is proposed. This method can guide large language models to extract knowledge from fault operation and maintenance data through zero-shot Prompts and complete the construction of large-scale fault operation and maintenance knowledge graphs iteratively. Experimental results show that the proposed method is scientific and effective.
    A Financial Knowledge Q&A Model for Power Enterprise Based on ChatGLM2-6B
    YE Xue, YANG Sheng, CHENG Kai, ZHU Feng
    2024, 0(11):  54-63.  doi:10.3969/j.issn.1006-2475.2024.11.009
    Asbtract ( 57 )   PDF (1481KB) ( 60 )  
    References | Related Articles | Metrics
    With the continuous expansion of the scale of the power system, a significant amount of repetitive and complex tasks emerge in daily financial operations. Traditional methods of organizing and managing financial knowledge are no longer sufficient to satisfy the requirements of the current power system. With the consideration of this, the paper constructs a financial knowledge graph using the large-scale language model called ChatGLM2-6B. This method aims to standardize financial and project management processes and assist in financial decision-making. Firstly, the ChatGLM2-6B model should be optimized through instruction fine-tuning and prompt learning in order to extract event and event relationship pairs from financial contracts and invoice data, respectively. Then, the event relationship pairs are then stored as a local knowledge base using the FAISS vector database, additionally, a FAISS-ERNIE similarity evaluation model is trained to enhance the capability of knowledge retrieval, which could improve the question-answering ability of ChatGLM2-6B. Finally, hierarchical clustering algorithm is employed to generalize event relationship pairs, aiming to obtain contract knowledge graph and invoice knowledge evolutionary graph. These two graphs could be utilized to provide standardized guidance and supervision for real-time financial operations, achieving transparency in financial execution. The experimental results demonstrate that the method proposed in this paper exhibits excellent performance in event extraction, event relationship pair extraction, and similarity retrieval. The constructed contract and invoice knowledge evolutionary graphs hold significant implications for financial management in power enterprises, contributing to enhance the level of corporate management. 
    Integrating Syntactic and Semantic Features for Automated Essay Scoring
    CHEN Yuhang1, YANG Yong1, Palidan TUERXUN1, FAN Xiaochao1, REN Ge1, DIAO Yufeng2
    2024, 0(11):  64-69.  doi:10.3969/j.issn.1006-2475.2024.11.010
    Asbtract ( 53 )   PDF (1175KB) ( 49 )  
    References | Related Articles | Metrics
     Automatic essay scoring is a technology that uses natural language processing technology to automatically evaluate and score essays. Automatic scoring of essays can improve scoring efficiency, reduce labor costs, ensure the objectivity and consistency of scoring, and has broad application prospects in the field of education. Although syntactic features and thematic features play an important role in automatic scoring of essays, so far, there is relatively insufficient research on how to better utilize these features for automatic scoring of essays. This paper proposes an automatic essay scoring method ISSF that integrates syntactic features and semantic features. The model uses Parser to extract the syntactic features of the essay, and uses BERT and adapter training methods to extract the deep semantic features of the essay. In order to better utilize the topic features and for the correlation between syntactic features and deep semantic features, the self-attention mechanism is used to extract thematic features of the essay and used to enhance syntactic features and deep semantic features. Experimental results show that the ISSF model proposed in this paper has achieved better average performance on 8 subsets of the public data set ASAP. Compared with baseline models such as Tongyi Qianwen, the ISSF model has a larger scoring range and complex scoring standards. In this case, it has more performance advantages.
    BiLSTM-Attention Prediction Model and Error Analysis #br# Based on Novel Multi-objective Coati Optimization Algorithm
    LI Junchao1, YOU Fei1, ZHANG Chao2, SU Lele2, GONG Yan2
    2024, 0(11):  70-76.  doi:10.3969/j.issn.1006-2475.2024.11.011
    Asbtract ( 47 )   PDF (1689KB) ( 46 )  
    References | Related Articles | Metrics
     Project cost prediction plays an important role in modern project management. However, due to market fluctuations, labor costs, and other factors, project cost forecasting has been challenging. Therefore, a novel multi-objective coati optimization algorithm is proposed, and a bidirectional long short-term memory network (BiLSTM) and attention mechanism optimized based on this algorithm are proposed to predict the cost of substation engineering. Firstly, the proposed algorithm is compared with the mainstream multi-objective optimization algorithm on 8 test problems, and the effectiveness of the multi-objective coati optimization algorithm is verified. Secondly, the proposed algorithm is used to optimize the prediction model to improve the accuracy of the model. The BiLSTM-Attention model captures the potential relationship in historical data to improve the accuracy and reliability of power transformation project cost prediction. Finally, the proposed model is compared with the five mainstream models, and the historical data of a 110 kV power transformation project in a province is used as a case study. The results show that the average absolute percentage error of the proposed model is 3.71%, which is reduced by 9.82 percentage points compared with BP, 5.81 percentage points compared with ANN, 5.40 percentage points compared with LSTM, 2.03 percentage points compared with LSTM-SVR, and 1.00 percentage points compared with CNN-LSTM.
    Network Public Opinion Prediction Based on Variational Mode Decomposition and IGJO-SVR
    ZHANG Zhixia, QIN Zhiyi
    2024, 0(11):  77-83.  doi:10.3969/j.issn.1006-2475.2024.11.012
    Asbtract ( 40 )   PDF (4288KB) ( 35 )  
    References | Related Articles | Metrics
     The prediction of the evolution trend of network public opinion has very important practical significance for the relevant government departments to supervise the development of public opinion and maintain the stability of public opinion in today's network environment. Aiming at the particularity of network public opinion data and considering the accuracy of model prediction results, this paper uses variational mode decomposition (VMD) and improved golden jackal optimization support vector regression (IGJO-SVR) to construct a network public opinion evolution trend prediction model, and takes ‘Beixi’ event-related public opinion data as a case for empirical research. The comparison results show that the accuracy of the prediction model constructed in this paper is significantly better than the other models. The network public opinion heat prediction model based on variational mode decomposition VMD and IGJO-SVR has excellent prediction accuracy, and can provide effective public opinion situation analysis and decision-making help for relevant government departments in practical work.
    Polyp Segmentation Based on Involution and Coordinate Reverse Attention 
    WAN Hongwei, CHEN Pinghua
    2024, 0(11):  84-90.  doi:10.3969/j.issn.1006-2475.2024.11.013
    Asbtract ( 45 )   PDF (1538KB) ( 50 )  
    References | Related Articles | Metrics
     Polyps in colon images are characterized by variable morphology and blurred edges. Aiming at the problems of the current neural networks for polyp segmentation, such as the inadequate feature extraction due to the inherent limitations of convolution, and the unsatisfactory segmentation due to the incomplete relationship between area and boundary, a network(IN-CRNet) based on Involution and coordinate reverse attention was proposed. In the encoder, an Involution-based Receptive Field Module(InRFB) was designed to adaptively capture contextual information at different scales. It improved the ability to detect complex and variable polyps. In the decoder, a coordinate reverse attention module(CRA) was designed to focus on the importance of both regions and edges and establish the relationship between them. It gradually refined the details of the edges from the bottom  to up. The experimental results on five public datasets show that IN-CRNet effectively improves the accuracy of segmentation and has good generalization ability. 
    YOLOLW: A Novel Lightweight Object Detection Model
    ZHANG Yu1, 2, LI Jing1, 2, MA Ming1, 2, WANG Zhongxiang1, 2, SUN Yan1, 2
    2024, 0(11):  91-98.  doi:10.3969/j.issn.1006-2475.2024.11.014
    Asbtract ( 75 )   PDF (2108KB) ( 71 )  
    References | Related Articles | Metrics
     In response to the growing demand for real-time mobile object detection deployment, the current YOLO backbone network falls short. Hence, this paper proposes YOLOLW, a lightweight object detection model based on the anchor frame. Firstly, it incorporates a novel lightweight decoupling header to enhance focus on classification and regression tasks and improve model accuracy. Secondly, it designs a lightweight and reparameterized network structure that achieves superior detection accuracy while maintaining its lightweight nature. Thirdly, it enhances the feature pyramid structure (FPN) by effectively integrating shallow features through dynamic convolution and cross-hierarchy association. Lastly, spatial and channel attention mechanisms are introduced to significantly boost the model’s accuracy. Experimental results validate the effectiveness of the YOLOLW model.
    Multi-scale Moving Object Detection Algorithm Based on Improved YOLOv7-tiny
    DONG Yuwen
    2024, 0(11):  99-105.  doi:10.3969/j.issn.1006-2475.2024.11.015
    Asbtract ( 59 )   PDF (3003KB) ( 56 )  
    References | Related Articles | Metrics
    Aiming at the problem of model misdetection and omission caused by the few pixels of long-distance intrusion object, insufficient texture information as well as large-scale transformations during the continuous approach of the object in regional security defence, an improved multi-scale moving object detection algorithm based on the YOLOv7-tiny algorithm is proposed. Firstly, a new OBM module is proposed for the feature extraction network, using a multi-dimensional attention mechanism to improve the feature extraction capability of the network. Secondly, an improved AC-BiFPN bidirectional feature fusion strategy is used to combine the multi-dimensional adaptive weighted fusion. The scale features are passed to the ACmix attention mechanism to improve the model’s perception of multi-scale objects. Finally, the activation function of the model is optimized to weight the area between the predicted frame and the real frame to reduce the model prediction bias. The model is tested on a self-made pedestrian and vehicle data set, and the experimental results show that compared with the original YOLOv7-tiny model, the improved YOLOv7-tiny model addresses the problem of misdetection and omission of pedestrians and vehicles during long-distances monitoring, with an increase of 3.96 percentage points in the detection accuracy, an increase of 2.22 percentage points in the average detection accuracy (mAP@0.5:0.95) , and the real-time frame rate reaches 32.7 fps on edge GPUs, which meets the practical application requirements.
    Multi-view 3D Reconstruction Based on Improved Self-attention Mechanism
    QI Xian, LIU Daming, CHANG Jiaxin
    2024, 0(11):  106-112.  doi:10.3969/j.issn.1006-2475.2024.11.016
    Asbtract ( 51 )   PDF (4076KB) ( 45 )  
    References | Related Articles | Metrics
    To address the current problems that multi-view 3D reconstruction cannot adapt to high-resolution scenes, poor completeness, and ignoring global background information, this paper proposes a 3D reconstruction network MVFSAM-CasMVSNet that fuses deformable convolution with improved self-attention mechanism. Firstly, a deformable convolution module dedicated to the task of multi-view stereo reconstruction is designed to adaptively adjust the range of extracted features and enhance the feature extraction capability for depth mutation. Secondly, considering the correlation of depth information and feature interactions among multiple views, a multi-view fusion self-attention module is designed to aggregate remote context information within each view by linear self-attention with low computational complexity, and capture the depth dependencies between the reference view and the source view by improved multi-head attention. Finally, the cost volume is constructed and regularized from coarse to fine using a multi-stage strategy, and depth map is generated using the cost volume with higher resolution. The test results on DTU dataset show that MVFSAM-CasMVSNet has respectively improved completeness, accuracy, and overall by 15.6%, 7.4%, and 11.8%, compared with baseline model, and has optimal overall compared with other existing models. Meanwhile, experimental results on the Tanks and Temples dataset show that the network has an average F-score improvement of 6.5% compared to the benchmark model. The method in this paper has excellent reconstruction effect and generalization ability for high-resolution scenes in the field of multi-view 3D reconstruction
    A River Discarded Bottles Detection Method Based on Improved YOLOv8 
    CHEN Kai1, LI Yiting1, 2, QUAN Huafeng1
    2024, 0(11):  113-120.  doi:10.3969/j.issn.1006-2475.2024.11.017
    Asbtract ( 64 )   PDF (3172KB) ( 49 )  
    References | Related Articles | Metrics
     Currently, object detection has been found extensive applications across various domains and is progressively reaching a state of maturity. However, in the task of detecting riverine discarded bottles, the ongoing challenges with existing image processing technologies include low accuracy, high costs, and deployment difficulties, which continue to be significant obstacles in this research field. Therefore, we present an enhanced deep learning detection model based on YOLOv8. Firstly, to tackle the common issue of false positives and false negatives for small objects in unmanned vessel images, we introduce the Bi-PAN-FPN concept to improve the Neck component of YOLOv8-n. By carefully considering and reusing multi-scale features, this approach aims to achieve a more sophisticated and comprehensive feature fusion process while minimizing parameter costs. Secondly, the CIoU loss function is optimized by introducing the EIoU and addressing the imbalance between hard and easy samples. This is achieved by separating the influence of aspect ratio factors between predicted and ground-truth boxes. These enhancements aim to improve the optimization capability of the model. We conduct experiments using the publicly available FloW dataset and includes multiple evaluations, such as ablation experiments, comparative experiments, performance analysis of loss functions, and contrast experiments in special scenarios. These evaluations provide comprehensive evidence of the feasibility and effectiveness of the proposed method from multiple perspectives. The results of the ablation experiments indicate that the enhanced YOLOv8-n model exhibits a significant improvement in reducing false negatives compared to the baseline network, achieving an average precision of 85.2%. This marks a 2.7 percentage points increase over the baseline network model, demonstrating a notable enhancement in detection performance. The results of the comparative experiments indicate that the average precision of the improved model surpasses six other models, namely Mobilenet-SSDv2, YOLOv4s, YOLOv5s, YOLOv7-tiny, YOLOv3-SPP, and YOLOv5-MobileNetV3s, by 60.15%, 18.99%, 3.90%, 7.30%, 28.7%, and 55.47%, respectively. These findings highlight the superior overall performance of the improved model in terms of FPS, parameter quantity, and model size. Additionally, the improved model demonstrates exceptional performance in special scenarios. Therefore, the comprehensive performance of the improved model in the paper outperforms the currently popular models for river garbage detection. As a result, it is more suitable for real-time detection of discarded bottles in rivers.
    Industrial Image Circle Detection Algorithm Based on Edge Drawing
    YANG Qingwu, LUO Xiaohui, LIU Xin
    2024, 0(11):  121-126.  doi:10.3969/j.issn.1006-2475.2024.11.018
    Asbtract ( 52 )   PDF (2366KB) ( 36 )  
    References | Related Articles | Metrics
    In this paper, a real-time, parameterless circle detection algorithm based on Hough transform is proposed to solve the problems of more parameters and more computation. This algorithm obtains several edge segments in the image by using ED (Edge Drawing) algorithm, and extracts the set of arc segments that may form a circle from the edge segment geometry by using line segment detector. Each pair of line segments is analyzed to determine whether a pair of effective line segments can be formed, thus generating an initial set of circles. Then the candidate circle is generated according to the geometric properties of the circle edge and verified by the reverse verification principle. The experimental results show that the proposed method can still detect the circle in the image under the interference of blur, shadow and occlusion, and the accuracy and stability of the industrial image are significantly improved compared with other algorithms.