Journal of Graphics

Table of Contents for Issue 1, 2022

2022, 43(1): 1-1.

Abstract ( 189 )

PDF (236KB) ( 126 )

Related Articles | Metrics

Visual analysis of origin-destination data

MA Xiao-dong, REN Peng-kun, ZHAO Fan

2022, 43(1): 1-10. DOI: 10.11996/JG.j.2095-302X.2022010001

Abstract ( 161 )

PDF (8172KB) ( 215 )

Origin-destination data was a kind of trajectory data composed of start point, end point, time, and some other attributes. It was a typical spatio-temporal data, which was generated in such fields as urban transportation management, population migration, and social media. Visual analysis technology was widely employed to study the spatio-temporal pattern of large-scale origin-destination data, accomplishing the deep exploration of data. Firstly, we introduced the characteristics of origin-destination data and the tasks of visual analysis. Secondly, we reviewed the existing visualization methods, interaction technologies, and visualization systems of origin-destination data in recent years, and presented the application cases in different fields. Finally, we summarized the problems and challenges in relevant research, and envisioned the prospect of the research on visualization of origin-destination data, shedding new light on future research.

Related Articles | Metrics

Information communications technology assisted tower crane safety management-review and prospect

DUAN Rui, DENG Hui, DENG Yi-chuan

2022, 43(1): 11-20. DOI: 10.11996/JG.j.2095-302X.2022010011

Abstract ( 138 )

PDF (1182KB) ( 88 )

Tower cranes are the most frequently used vertical transportation tools in construction projects. Due to the high operating risks of tower cranes, it remains urgent to improve its safety management level in the industry. In recent years, information and communication technology (ICT) has gradually been applied to the safety management of tower cranes, but the current applications in this field are mostly single-point development, and the level of intelligence is low, which fails to meet the existing safety management requirements of tower cranes. By investigating the relevant literature in the field of tower crane safety management in the past decade, the current research status of tower crane safety management by domestic and foreign scholars was sorted out and summarized, and analyses were made on the object, core technology, attention, and advantages and disadvantages of the existing research. In addition, through the analysis of the existing methods, objectives and framework were proposed for the future tower crane safety management. The potential building information modeling (BIM) and computer vision (CV) in ICT were integrated into the tower crane safety management framework, so as to realize the real-time monitoring and early risk-warning of the tower crane operation process, and to effectively reduce the occurrence of accidents. The proposed framework is expected to promote the intelligentization and informatization of the tower crane safety management, and to provide some inspiration for the healthy development of construction engineering.

Related Articles | Metrics

PCB defect detection based on convolutional neural network

HE Guo-zhong, LIANG Yu

2022, 43(1): 21-27. DOI: 10.11996/JG.j.2095-302X.2022010021

Abstract ( 584 )

PDF (8002KB) ( 478 )

In the production of printed circuit boards (PCB), the production process and other problems incur flaws and defects on the circuit board. In order to enhance the detection efficiency of circuit board defects, a circuit board defect detection network based on convolutional neural network (CNN) was proposed. The whole detection network was optimized and reconstructed based on the YOLO v4 network. Aiming at the difficulty of precise and complex PCB production and difficult detection of various defects, a long-distance global attention mechanism based on fine-grained spatial domain was added to the optimized network. At the same time, on the basis of the spatial pyramid pooling (SPP) module, the feature map was reorganized as the input of each YOLO detection head. The long-distance attention mechanism channel was adopted to transfer the features extracted from the shallow network to the deep network, and the feature map reorganization method was utilized to boost the richness of feature information, thereby improving the accuracy of PCB defect detection. After experimental analysis, compared with various classic convolutional neural networks, the proposed algorithm is greatly superior in PCB board defect detection tasks. The mean average precision (mAP) of the overall defects reaches 91.40%, which is suitable for actual production and testing links.

Related Articles | Metrics

Solanaceae disease recognition method based on capsule SE-Inception

YANG Yong-bo , ZHAO Yuan-yang , LI Zhen-bo, LI Ye

2022, 43(1): 28-35. DOI: 10.11996/JG.j.2095-302X.2022010028

Abstract ( 140 )

PDF (1502KB) ( 101 )

Aiming at the diseases of two types of Solanaceae vegetables, tomato and eggplant, a noise-resistant Solanaceae disease identification network was constructed based on SE-Inception and capsule network, called capsule SE-Inception. The network is mainly divided into two parts: the feature extraction part and the capsule network part. The feature extraction part of the network employed a batch normalization layer (BN) to accelerate the convergence of the network; the SE-Inception structure and multi-scale feature extraction module were used to improve the accuracy of the model. The capsule network part utilized a capsule with a routing iteration number of two and a dimension of sixteen for processing. The experiments were undertaken based on a self-built data set of Solanaceae diseases. Our sample data contains four disease categories: whitefly, powdery mildew, yellow smut, and cotton blight, as well as healthy leaves. Besides, in order to reduce over-fitting, the data was augmented. The experimental results show that the capsule SE-Inception network displays good noise immunity against common Gaussian, salt and pepper, and fuzzy noise; it only needs a limited amount of data to achieve higher recognition accuracy. Based on the same amount of data, the recognition accuracy of capsule SE-Inception network outperforms that of common lightweight models.

Related Articles | Metrics

Adaptive multi-class centers and semi-heterogeneous network for sketch-based 3D model retrieval

BAI Jing, TUO Ji-wen , BAI Shao-jin , YANG Zhan-yuan

2022, 43(1): 36-43. DOI: 10.11996/JG.j.2095-302X.2022010036

Abstract ( 228 )

PDF (2354KB) ( 152 )

Sketches are advantageous in being easy to construct and unrestricted by language, discipline, age, and so forth, and the 3D model retrieval based on hand-drawn sketches has attracted increasing attention. However, due to the complexity of 3D models, intra-class diversity of 2D sketches, and the inter-domain differences between 3D models and 2D sketches, the sketch-based 3D model retrieval remains highly challenging currently. To address these issues, we proposed a 3D model retrieval for sketch based on adaptive multi-class centers and semi-heterogeneous network. First, the initial features of the sketches and the 3D models were extracted separately through two heterogeneous networks: a sketch feature embedding sub-network based on adaptive multi-class centers was designed to capture the intra-class diversity of sketches, and a 3D model feature embedding sub-network based on multi-view feature fusion was adopted to adapt to the complexity of 3D models. Then, using the label vectors with rich semantic information as guides, a homogeneous network was designed to realize the cross-domain shared feature embedding of the sketches and 3D models, so as to reduce the inter-domain differences. Comparative experiments on the large public data sets SHREC2013 and SHREC2014 demonstrate that the proposed algorithm is on par with or better than the state-of-the-art methods.

Related Articles | Metrics

Acquisition method of specific motion frame based on human attitude estimation and clustering

CAI Min-min, HUANG Ji-feng, LIN Xiao, ZHOU Xiao-ping

2022, 43(1): 44-52. DOI: 10.11996/JG.j.2095-302X.2022010044

Abstract ( 161 )

PDF (3617KB) ( 171 )

The acquisition of specific motion frames in motion video was an important part of intelligent teaching. In order to obtain specific motion frames in video for further analysis, a method of extracting specific motion frames from motion video was proposed using the knowledge of pose estimation and clustering. Firstly, the HRNet attitude estimation model was adopted as the basis, which was of high precision but large scale. To meet the needs of practical application, this paper proposed a Small-HRNet network model by combining it with the data encoding of DARK. The parameters were reduced by 82.0% while the precision was kept unchanged. Then, the Small-HRNet model was employed to extract human joint points from the video. The human skeleton feature in each video frame served as the sample point of clustering, and finally the whole video was clustered by the skeleton feature of the standard motion frame as the clustering center to produce the specific motion frame of the video. The experiment was carried out on the martial arts data set, and the accuracy rate of the martial arts action frame extraction was 87.5%, which can effectively extract the martial arts action frame.

Related Articles | Metrics

Video super-resolution reconstruction based on multi-scale time domain 3D convolution

TANG Xiao-tian, MA Jun , LI Feng , YANG Xue , LIANG Liang

2022, 43(1): 53-59. DOI: 10.11996/JG.j.2095-302X.2022010053

Abstract ( 199 )

PDF (4639KB) ( 146 )

Video super-resolution was a work of great practical value. In view of the lack of high-resolution resources in the ultra-high-definition industry, to efficiently utilize the rich temporal correlation information and spatial information between video sequence frames, a video super-resolution reconstruction algorithm based on multi-scale time-domain 3D convolution was proposed. The algorithm extracted the spatiotemporal features of the input low-resolution video sequence frames through the 3D convolution of different time scales. 3D convolution can simultaneously model space and time, which is more suitable for processing video tasks than 2D convolution. After the adaptive motion compensation of two spatio-temporal features extracted in different scales and time domains, the sub-pixel convolutional layer performed resolution enhancement, which was added to the up-sampled input frame to obtain the final reconstructed high-resolution image. The experimental results on the standard data set show that the algorithm can significantly boost visual effects and objective quality evaluation indicators such as peak signal-to-noise ratio and structural similarity, outperforming algorithms such as FSRCNN and EDSR.

Related Articles | Metrics

Object tracking of reverse joint sparse representation with local template update

YU Hong-ling, CHEN Ying-pin , XU Yan-ping , LIN Chen , JIANG Min-yi , LUO Cong-miao , CHEN Yue , LIN Yao-jin,

2022, 43(1): 60-69. DOI: 10.11996/JG.j.2095-302X.2022010060

Abstract ( 170 )

PDF (13561KB) ( 181 )

The reverse joint sparse representation algorithm can make full use of the temporal similarity and spatial continuity in the tracking process. However, tracking drift can be easily incurred under the influence of occlusion and illumination change. Aiming at this problem, we proposed the reverse joint sparse representation tracker (RJST). It can accomplish the reverse joint sparse representation through the reversely local reconstruction of the object template set. Firstly, the object template set was initialized in the first frame, and the candidate images were generated by particle filtering. They were partitioned into blocks, and the reverse joint sparse representation model was constructed. Then, the sparse coding matrix was solved using the alternating direction method of multipliers. The optimal candidate image was acquired by the two-step scoring mechanism. Finally, whether the current object had local occlusion was evaluated according to the similarity score. If there was no occlusion, the object template set was locally updated to eliminate the tracking drift. Experimental results show that the precision and success rate of RJST reached 85.4% and 62.8% on the OTB-2013 benchmark, and 76.8% and 68.6% on the OTB100 benchmark, respectively, and that the speed was 5.76 frames per second, which can effectively boost robustness and eliminate tracking drift.

Related Articles | Metrics

Blind image inpainting based on context gated convolution

WEN Jing, DING You-dong, YU Bing

2022, 43(1): 70-78. DOI: 10.11996/JG.j.2095-302X.2022010070

Abstract ( 218 )

PDF (2086KB) ( 153 )

Image inpainting methods based on deep learning have achieved great progress. At present, most of the image inpainting methods use the input mask to reconstruct the degraded areas of the image. Based on this observation, a two-stage blind image inpainting network was proposed, comprising a mask prediction network and an image inpainting network. The input of a mask was not required in the whole inpainting process. The mask prediction network could automatically detect the degraded area of the image and generate a mask according to the input image, and the image inpainting network could restore the missing part of the input image based on the prediction mask. In order to make better use of global context information, a context-gated residual block (CGRB) module was designed based on context-gated convolution to extract feature information. In addition, the spatial attention residual block (SARB) was proposed to model the relationship between pixels in the long-distance image, filtering some irrelevant details. A large number of experimental results on the CelebA-HQ, FFHQ, and PairsStreet datasets show that the improved algorithm is superior to other comparison methods and can generate convincing images.

Related Articles | Metrics

Isogeometric analysis of bending strength of spur gear

XUE Yu-tong, ZHAO Gang, WANG Ai-zeng, HE Chuan

2022, 43(1): 79-84. DOI: 10.11996/JG.j.2095-302X.2022010079

Abstract ( 117 )

PDF (845KB) ( 122 )

As an important mechanical part, gear is widely applied in all kinds of mechanical equipment, and its design and manufacture directly affect the actual performance and work quality of mechanical equipment. Since the working life of the gear is inversely proportional to the sixth power of the maximum bending stress value of the gear, accurate calculation of the bending strength of the gear root is a necessary guarantee to extend its service life. In order to address the bending stress of the gear more accurately, this paper proposed an algorithm of gear bending strength analysis based on isogeometric analysis method, which provided the isogeometric analysis of the mechanical properties for the two-dimensional gear structure. Compared with the traditional calculation formula of tooth root bending stress and the finite element method, our results show that the isogeometric analysis approach is of higher accuracy and efficiency in analyzing the bending stress of the gear tooth root. Compared with the stress field obtained by the finite element method, the result by the isogeometric analysis is smoother, which provides an effective method for solving the bending strength problem of gear root.

Related Articles | Metrics

Real time outdoor shadow detection technology for mobile augmented reality

QIU Dong, WU Yun-chao, LI Wei-qing, SU Zhi-yong

2022, 43(1): 85-92. DOI: 10.11996/JG.j.2095-302X.2022010085

Abstract ( 151 )

PDF (24419KB) ( 129 )

Aiming at the problems of high false detection rate and poor edge continuity in current video shadow detection algorithms in mobile view, a real-time shadow detection algorithm based on track and detection framework was proposed. Firstly, the overlapped shadow parts of the two frames were tracked twice, the tracking points with larger error were filtered by forward and backward tracking, and the accuracy of the tracking edge was ensured by Canny edge confidence. Then, the new region to be detected was obtained by the region division method based on optical flow. Secondly, seven-dimensional feature vectors were constructed for texture edge error detection, soft shadow detection, and dark area error detection. Then the support vector machine (SVM) classifier was trained by extracting feature vectors from shadow edge, and the trained classifier was employed to detect the shadow in the new area. Finally, for the broken edges in the detection results, an algorithm based on RGB color space gradient direction consistency was proposed to optimize the connection of the broken edge. Experimental results show that the proposed algorithm exhibits the best comprehensive performance compared with the latest research results, and is superior to the existing methods in terms of detection accuracy and edge continuity.

Related Articles | Metrics

MvUPA: universal perturbation attack against 3D shape retrieval based on multi-view networks

TANG Jing, PENG Wei-long, TANG Ke-ke, FANG Mei-e

2022, 43(1): 93-100. DOI: 10.11996/JG.j.2095-302X.2022010093

Abstract ( 109 )

PDF (1353KB) ( 77 )

Geometric deep learning models have been applied in 3D shape retrieval task successfully, and their security evaluation is also drawing the attention of researchers. This paper proposed a method of multi-view universal perturbation attack (MvUPA) for 3D shape retrieval evaluation, so as to generate perturbed samples with a higher success rate of attack. Firstly, a multi-view depth panoramic map-based network was designed to train an efficient embedding representation for multi-view 3D shape retrieval. Secondly, a fusion loss function and its attack mechanism beneficial to multi-input UPA was proposed. The loss function combined triplet loss and label loss, thereby improving the perturbation generation for different categories of samples with similar topology and same category samples with different topology. The experiments validated the attack effectiveness and stability of MvUPA on multi-view retrieval models. MvUPA brought the decrease rate (DR) up to 94.52%, and the DR of the fused loss function was about 3.0%–5.5% higher than that of a single loss function.

Related Articles | Metrics

Graph convolution network based BREP→CSG conversion method and its application

ZHOU Bo, GUO Zheng-yue, HAN Cheng-cun, DU Hua, YAN Yi-man, LUO Yue-tong

2022, 43(1): 101-109. DOI: 10.11996/JG.j.2095-302X.2022010101

Abstract ( 182 )

PDF (1136KB) ( 117 )

Boundary representation (BREP) and construction solid geometry (CSG) serve as the two most widely employed entity representations. There remains an urgent need for the BREP→CSG automatic conversion algorithm in such fields as particle transport calculation auxiliary modeling. However, the most commonly adopted segmentation-based BREP→CSG conversion algorithm is disadvantageous in “large amount of calculation and too complicated CSG expression”. Through the observation that “the CSG expression structure of the topologically similar BREP model is similar”, it was proposed to establish a model library containing the two tuples BREP and CSG. For the BREP model to be converted, the similar model was retrieved from the model library, and then the conversion result was generated based on the CSG expression of the similar model. On the one hand, this method can improve the conversion speed, and on the other hand, by optimizing the CSG expression, it can overcome the shortcomings of the space-based segmentation method. The extended attribute adjacency graph was applied to the description of the topological characteristics of the BREP model, the model similarity problem was regarded as the attribute adjacency graph classification problem, and then the graph convolutional network (GCN) was utilized to achieve fast model retrieval. The extended attributes of the attribute adjacency graph were also carefully designed to boost the accuracy of model retrieval. The algorithm has been integrated into the self-developed particle transport visual modeling software cosVMPT (COSINE visual modelling of particle transport), and tests were performed using the typical complex component divertor model in China Fusion Engineering Test Reactor (CFETR). The test results show the time validity of the algorithm and the superiority of the CSG results.

Related Articles | Metrics

An adaptive isogeometric collocation method with improved PHT-splines

JIA Yue, ANITESCU Cosmin, LI Chun

2022, 43(1): 110-117. DOI: 10.11996/JG.j.2095-302X.2022010110

Abstract ( 114 )

PDF (2289KB) ( 89 )

The Gaussian isogeometric analysis (IGA) collocation method was extended to arbitrary higher order polynomial degrees. The current IGA collocation method applied a new hierarchical basis over T-meshes (PHT-splines), which took advantage of the tensor product structure to prevent the decay phenomenon from happening in the original PHT basis. The improved method collocated at Gaussian points as the superconvergent points for the new PHT elements. Based on the new PHT basis, the current collocation method can be extended to arbitrary higher order approximation. In order to simplify the collocation boundary condition, a hybrid method was adopted to impose the boundary condition, using the Galerkin method for the boundary part and combing with the collocation solving system. The local refinement strategy was driven by a recovery-based error estimator that invoked computing an improved approximation without knowledge of the exact solution. The proposed collocation method can obtain the optimal convergent rates, compared with the IGA Galerkin method.

Related Articles | Metrics

Guided normal GPU filtering of depth images

CHONG Si-jie, WANG Shi-wei, LIU Li-gang

2022, 43(1): 118-124. DOI: 10.11996/JG.j.2095-302X.2022010118

Abstract ( 103 )

PDF (9292KB) ( 156 )

Depth images acquired by depth cameras generally contain noises and lose detailed geometric information. Thus, the filtering of depth images has become an important topic in both computer graphics and computer vision. However, most current filtering methods can hardly preserve the sharp features in the objects and often result in over-smoothing results. To this end, we proposed a novel joint bilateral filtering method for filtering depth images. First, we estimated the normal of each pixel in the depth image. Then we computed the weight of the normals by voting to perform joint bilateral filtering on all pixels. Finally, the vertex coordinates were updated according to the filtered normals. This method took into account the texture information with high accuracy as guidance information, which can yield more reliable filtering effects. In addition, this method was based on the local information of the point cloud, did not need to solve large matrixes, and employed GPU parallelism leading to extremely high computational efficiency. Experiments show that our method can highly preserve the edges in the normal field, thus preserving sharp features better than previous methods.

Related Articles | Metrics

Simulation and prediction of regional pollutants based on INLA-SPDE method

YUAN Ze, CHEN Bin

2022, 43(1): 125-132. DOI: 10.11996/JG.j.2095-302X.2022010125

Abstract ( 121 )

PDF (2185KB) ( 58 )

The simulation and prediction of regional pollutants generally use the traditional spatial interpolation method, which cannot obtain accurate results when the source data is not uniformly distributed. To address these problems, a method for simulation and prediction of regional pollutants based on the INLA-SPDE model was proposed. The interpolation model was based on a Bayesian hierarchical model where the spatial-component was represented through the stochastic partial differential equation (SPDE) approach, with a lag-1 temporal autoregressive component (AR1). In addition, the model included 10 spatial and spatio-temporal predictors such as meteorological variables. By building 12 models for each month with the integrated nested Laplace approximation (INLA), this research realized the spatio-temporal simulation and prediction of PM2.5 concentration at daily resolution in the Beijing-Tianjin-Hebei region in 2019. Experiments show that compared with traditional Kriging interpolation methods, the proposed model can yield a better prediction of air pollutants at regional scale. Particularly, the prediction accuracy of high-value pollutants was improved significantly, and air pollutants exceedance probabilities can also be generated. Furthermore, a system for regional PM2.5 concentration simulation and decision support was established, the system can provide support for the travel of ordinary people or the decision-making of government officials, and verify the practicability and value of the proposed model.

Related Articles | Metrics

Future-frame-based temporal anti-aliasing research and practice

DU Xing-sheng, WU Tong, ZHANG Jing-yi, LI Gen, LI Xin, ZHANG Yan-ci

2022, 43(1): 133-140. DOI: 10.11996/JG.j.2095-302X.2022010133

Abstract ( 145 )

PDF (3587KB) ( 97 )

Highly advantageous in efficiency, the algorithm of temporal anti-aliasing has been one of the most widely employed real-time anti-aliasing algorithms in recent years. To achieve real-time anti-aliasing, this algorithm assigns the sampling points to the multiple history frames and reuses historical data. When the sampling information in the time domain is sufficient and historical data is usable, it can achieve a similar effect as supersampling anti-aliasing can. However, problems such as geometric edge jagging, ghosting, and subpixel detail missing will arise in the use of this algorithm in practical applications. Based on the research of temporal anti-aliasing, this thesis proposed the future-frame-based temporal anti-aliasing to solve the problems. The basic idea of the proposed algorithm was that: using the existing information in the time domain, it took the next aliasing future frame into account, using samples in future frame to enhance the geometric anti-aliasing effect, achieve anti-ghosting, and improve the situation of subpixel detail missing. In the process of implementation, this thesis proposed solutions to reusable future data extraction and future data reuse. Experimental results indicate that compared with the temporal anti-aliasing, the proposed algorithm can yield a better effect.

Related Articles | Metrics

Intelligent inspection and maintenance of mechanical and electrical equipment based on MR

WANG Wei, HONG Xue-feng, LEI Song-gui

2022, 43(1): 141-148. DOI: 10.11996/JG.j.2095-302X.2022010141

Abstract ( 93 )

PDF (3664KB) ( 81 )

This paper examined the basic structure and training method of Faster R-CNN (convolutional neural networks) target detection network. The state data set of mechanical and electrical equipment was established, and the target detection network was trained. In a single step, the region of pointer instrument could be extracted, and the reading of digital instrument and the state of switch and plug could be recognized. The target detection network was tested under different viewing angles and illumination intensities. The results show that the model can maintain the accuracy of more than 90% in different environments. Finally, based on the reasoning results, the intelligent maintenance assistant system for mechanical and electrical equipment developed based on Unity 3D software and HoloLens 2 hardware was applied to the retrieval of the mixed reality (MR) holographic induction maintenance information, thus guiding the operation of the support personnel. In order to verify the availability of the system, the experimental verification process was added, and the experimental results show that the experimenter could complete the maintenance task quickly and efficiently using MR. In addition, test and evaluation were conducted based on the operation time and questionnaire survey, and qualitative analysis was carried out regarding the advantages of the system.

Related Articles | Metrics

Partitioned scanning path planning for selective laser melting of thin-walled parts

DENG Yang-yang, LI Wei-shi

2022, 43(1): 149-155. DOI: 10.11996/JG.j.2095-302X.2022010149

Abstract ( 104 )

PDF (1277KB) ( 120 )

Residual stresses in thin-walled parts manufactured with selective laser melting can lead to deformation and even cracking in the part. A partitioned scanning path planning strategy based on medial axis transformation was proposed to reduce the residual stress and control the distribution. Firstly, the medial axis transformation of the layered solid area was computed, and then the layered solid area was divided into main areas and connecting areas, according to the pruned medial axis. A sinusoidal scanning path and a regular triangle mesh scanning path were designed for the main area and the connecting area, respectively. Finally, the result of a typical multi-connected and thin-walled model was yielded.

Related Articles | Metrics

Location of cast-in-place concrete structural members based on BIM + CV

XUE Jing-guo, HOU Xue-liang

2022, 43(1): 156-162. DOI: 10.11996/JG.j.2095-302X.2022010156

Abstract ( 211 )

PDF (3268KB) ( 294 )

The positioning of building structural components is the key to automatic progress tracking. In the existing research results, the monitoring object is a simple construction site, and the form of today’s buildings is increasingly complex and is mainly of cast-in-place concrete structure. In order to track the construction progress of this kind of building, based on building information modeling (BIM) and computer vision (CV) technology, a method was proposed to locate the structural members of cast-in-place concrete building from the top view. Firstly, target detection technology was employed to identify the structural components in the construction state from the top view of the construction site. The BIM elements were projected onto the imaging plane based on the principle of camera imaging. Finally, the mapping relationship between the actual components and BIM elements was established through the results of registration and projection, so as to realize the automatic positioning of structural components. The experimental results show that the recognition accuracy is 82.08% and the registration rate is 94.17%. The recognition and registration effect of the plane formwork is the best. Compared with the similar methods, this method reduces the degree of occlusion, makes full use of the context information, and is more in line with the actual construction process. It provides a new scheme for the construction process tracking of complex buildings.

Related Articles | Metrics

Research on semantic 3D building modeling with multiple levels of detail

ZHANG Wen-yuan, TAN Guo-xin

2022, 43(1): 163-171. DOI: 10.11996/JG.j.2095-302X.2022010163

Abstract ( 326 )

PDF (7502KB) ( 222 )

Most of the existing three-dimensional (3D) modeling methods focus on geometric model construction, which neglect the semantic representation and multiple levels of detail, making the outcomes difficult for data interoperability. To address these issues, this paper proposed a set of 3D modeling techniques to generate geometric and semantic coherent building models on the basis of city geography markup language (CityGML), an open international standard for 3D virtual data storage and exchange. Geometry extrusion, semantic information extraction based on geometry features and rules, geometric and semantic mapping were explored for generating CityGML models with multiple levels of detail. A WebGL-based 3D GIS platform was developed utilizing the open-source Cesium library, and CityGML building models with four levels of detail were presented on the case study of campus area of the Hong Kong Polytechnic University. Experimental results indicate that the proposed modeling approach is able to make full use of the existing 2D building footprints, 3D geometric models, and building information models, and that the generated CityGML models with geometric, semantic, topological and appearance characteristics can provide better solutions for various applications such as 3D building model interoperability, data sharing, spatial analysis, and fine management.

Related Articles | Metrics

Work safety and interaction design strategies of hospital disinfection robot

WANG Qiu-hui, WANG Ya-xin

2022, 43(1): 172-180. DOI: 10.11996/JG.j.2095-302X.2022010172

Abstract ( 180 )

PDF (1896KB) ( 218 )

Safety and design quality play a key role in enhancing the operation efficiency of “Human-Robot System” of hospital public environment disinfection robots. Based on ergonomics, interaction science, and cognitive science, explorations were conducted on the safety design and interaction strategies of the hospital disinfection robot. The hierarchical task method was employed and combined with the table task method (HTA-T) to construct the task flow of disinfection operation. The typical safety accidents in the operation process and causative logic were analyzed, thereby establishing an abstract model of disinfection tasks based on the work domain analysis (WDA) method. The SRK decision analysis method was utilized to judge the cognitive behavior pattern mode (skill-based, rule-based, or knowledge-based) of hospital disinfection robot operation, and to match the relationship of cognitive factors. The theoretical model framework of operation safety design evaluation for the disinfection robot was constructed, and the design strategies based on four safety indicators were proposed: functional requirements, morphological structure, appearance color, and human-robot interaction. The outcome can shed light on the basic theoretical research and design practice of human-robot ergonomics of hospital public environment disinfection robots.

Related Articles | Metrics

Published as 1, 2022

2022, 43(1): 181-181.

Abstract ( 68 )

PDF (101578KB) ( 319 )

Related Articles | Metrics