Table of Contents

31 December 2020, Volume 41 Issue 6 Previous Issue    Next Issue

---

For Selected:

Download Citations EndNote Ris BibTeX

Toggle Thumbnails

Select

Review and analysis of image defogging algorithm  WANG Dao-lei, ZHANG Tian-yu 2020, 41(6): 861-870.  DOI: 10.11996/JG.j.2095-302X.2020060861 Abstract ( 900 )   PDF (6604KB) ( 579 )   Abstract: Image defogging is an image preprocessing method for man-machine recognition by analyzing and preprocessing the image with fog, meeting the application requirements under specific conditions. The influence of haze could incur lost details and low contrast for the image taken in foggy conditions, which would impact the subsequent analysis and recognition of the image. The past research on image defogging algorithms was summarized, such as image enhancement, image restoration, convolution neural network, and the improved algorithms, some of which were tested, evaluated, and analyzed in terms of advantages and disadvantages. Explorations were made on the future development, and suggestions were propounded for the difficult and easy parts of the algorithm, thus boosting the further development of the image defogging algorithms. Related Articles | Metrics

Select

Research progress and prospect of machine vision technology  ZHU Yun, LING Zhi-gang, ZHANG Yu-qiang 2020, 41(6): 871-890.  DOI: 10.11996/JG.j.2095-302X.2020060871 Abstract ( 3359 )   PDF (914KB) ( 1757 )   Abstract: Developed from the engineering of computer vision theory, machine vision involves such technologies as optical imaging, visual information processing, artificial intelligence, and mechatronics. With the transformation and upgrading of China’s manufacturing industry and development of relevant research, the machine vision technology, with the advantages of high precision, real-time performance, high-level automation and intelligence, has become one of the most significant driving forces for enhancing the intelligence of robots, thus being widely applied in modern industrial, agricultural, and military fields. In order to provide some guidance for researchers and engineers, this paper summarized an abundance of literature on machine vision technology, of which the research, development, and application in the recent decade were analyzed. Firstly, the development and current situation of machine vision around the world was introduced. Secondly, explorations were made on the key components of machine vision system including lighting, optical lenses, and cameras, vision processing algorithms including image preprocessing, image visual position, and segmentation, and the mainstream industrial software of machine vision. Thirdly, four typical applications of machine vision technology were presented, including production defect detection, intelligent visual surveillance, autopilot and assisted driving, and medical imaging diagnosis. Finally, the current challenges faced by machine vision were analyzed, with the future trends of machine vision predicted. Thus, this paper will play an active role in the development and application promotion of machine vision science and technology. Related Articles | Metrics

Select

An analysis of occlusion influence on object detection ZHANG Sheng-hu , MA Hui-min 2020, 41(6): 891-896.  DOI: 10.11996/JG.j.2095-302X.2020060891 Abstract ( 290 )   PDF (1335KB) ( 179 )   Abstract: The occlusion problem poses challenges to the current object detection. The presence of occlusion could destroy the overall structure of the object, which is likely to incur missing detections and false positives during the detection. Although the common methods for handling occlusion have greatly enhanced the performance of occlusion detection, there remains no specific quantitative analysis of the occlusion components and the impact of different occlusion ratios on the detection performance. In this paper, based on the data-driven method, a large number of uniform occlusion datasets were generated by simulation, named as More than Common Object Detection (MOCOD), and the detection performance under different occlusion ratios was analyzed quantitatively. On the basis of the analysis of occlusion’s influence, according to the occlusion ratios, the decay weight was introduced to select high-quality positive samples for the model training, thereby effectively improving the detection performance under occlusion conditions. Related Articles | Metrics

Select

DenseNet-attention for hyperspectral remote sensing image classification   ZHANG Yong-peng, ZHANG Chun-mei, BAI Jing 2020, 41(6): 897-904.  DOI: 10.11996/JG.j.2095-302X.2020060897 Abstract ( 226 )   PDF (1769KB) ( 253 )   Abstract: A new neural network, called DenseNet-Attention (DANet), was proposed in this paper for hyperspectral images classification to solve the problems of small sample quantity, insufficient features extraction, and indiscriminating contribution of the extracted features. First, it employed the three-dimensional convolution kernel to simultaneously extract both spectral and spatial features. Meanwhile, due to its dense blocks, DenseNet can not only fully extract more robust features, but reduce a large number of parameters. Second, the self-attention mechanism was added to the dense block as a module. Before the extracted feature was passed into the next layer of network, the weight was assigned to the feature according to its contribution through this model, thus strengthening the representation of the feature with ground object information. DANet was an end-to-end deep learning framework, which took the neighborhood block of the original hyperspectral image as an input without any preprocessing. Comparative experiments on Indian Pines and Pavia University datasets show that the classification accuracy of the network model proposed in this paper can reach 99.43% and 99.99% respectively, effectively enhancing the classification accuracy of hyperspectral images. Related Articles | Metrics

Select

CV image segmentation model combining with local and global features of the target  LI Xiao-hui, WANG Xi-li 2020, 41(6): 905-916.  DOI: 10.11996/JG.j.2095-302X.2020060905 Abstract ( 64 )   PDF (1354KB) ( 49 )   Abstract: With the development of the remote sensing satellite technology, high-resolution remote sensing images are on an increasing trend. The automatic target extraction from remote sensing images containing other information and complex background urgently needs to be realized. The traditional image segmentation method mainly depended on such underlying features as image spectrum and texture, and in image segmentation tasks, was likely to be impacted by the interference of shadow and occlusion in the image, complicating the segmentation and leading to unsatisfactory results. For this reason, according to the specific target type, a CV (Chan Vest) image segmentation model combined with local and global features of the target was proposed. Firstly, the deep learning generation model-CRBM (convolution restricted Boltzmann machine) was employed to represent the global shape features of the target and to reconstruct the shape of the target. Secondly, the edge information of the target was extracted by Canny operator, and a new shape constraint term integrating the local edge and global shape information was obtained by symbolic distance transformation. Finally, the CV model served as the image target segmentation model, and new constraints were added to gain the CV remote sensing image segmentation model integrating the local and global features of the target. The experimental results on the remote sensing dataset Levir-oil drum, Levir-ship and Levir-airplane show that the proposed model can not only overcome the noise sensitivity of the CV model, but also segment the target completely and accurately in the case of limited training data, small target size, occlusion and complex background. Related Articles | Metrics

Select

Plaintext correlation image encryption algorithm based on hyper-chaotic system   RUI Jie, HANG Hou-jun 2020, 41(6): 917-921.  DOI: 10.11996/JG.j.2095-302X.2020060917 Abstract ( 122 )   PDF (772KB) ( 111 )   Abstract: With the rapid development of computer networks and multimedia technologies, the security problem of digital image transmission has become increasingly salient. A hyper-chaotic encryption algorithm was proposed for plaintext associated images. The algorithm employed hyper-chaotic systems with more complex dynamics as a chaotic sequence generator, expanded the key space, and utilized a diffusion-scrambling-diffusion encryption framework. The method performed three-stage processing on the image, of which the depth diffusion improved efficiency using fewer iterations. The scrambling algorithm eliminated the correlation between adjacent pixels in the original image by disturbing the pixel positions. The experiment analyzed and compared the key space size, algorithm efficiency, and anti-plaintext attack capability. The simulation results show that the algorithm has not only larger key space and higher operating efficiency, but also strong plaintext sensitivity, thus effectively resisting differential attacks with great potential for secure communication applications. Related Articles | Metrics

Select

FANET: light field depth estimation with multi-channel information fusion  HE Ye, ZHANG Xu-dong, WU Di 2020, 41(6): 922-929.  DOI: 10.11996/JG.j.2095-302X.2020060922 Abstract ( 75 )   PDF (2630KB) ( 97 )   Abstract: Compared with the traditional two-dimensional images, the images, generated by the light field camera capturing the spatial and angular information of the scene in only one shot, contain more information and exhibit more advantages in the depth estimation task. In order to obtain high-quality scene depth using light field images, a feature assigning network, of which the structure can efficiently fuse the multi-channel information, was designed for depth estimation based on its multi-angle representation. On the basis of the artificial selection of specific views, convolution kernels of different sizes were utilized to cope with different baseline changes. Meanwhile, a feature fusion module was established based on the multi-input characteristics of light field data, and the double-channel network structure was used to integrate the front and back layer information, boosting the learning efficiency and performance of the network. Experimental results on the new HCI data set show that the network converges faster on the training set and can achieve accurate depth estimation in non-Lambertian scenes, and that the average performance on the MSE indicator is superior to other advanced methods. Related Articles | Metrics

Select

Extraction of mural paint loss regions based on spectral dimensionality reduction and Hu moment CAO Peng-hui, LYU Shu-qiang, WANG Wan-fu , GAO Zhen-hua, HOU Miao-le, 2020, 41(6): 930-938.  DOI: 10.11996/JG.j.2095-302X.2020060930 Abstract ( 82 )   PDF (7291KB) ( 77 )   Abstract: The extraction of mural paint loss plays an important role in investigating the present situation of murals. Given the similarity of spectral features of paint loss to the white patterns of the mural, the only utilization of spectral features would make it less accurate for the extraction of the mural paint loss. Aiming at improving the extraction performance, a comprehensive method was proposed that integrated spectral features and Hu moment. First, the supervised support vector machine method was employed to classify paint loss and white patterns by the spectral features which the dimension had been reduced. The classified pixels of paint loss and white patterns were then connected to form two types of polygons, which were regarded as the smallest indivisible objects. Subsequently, seven Hu moments for each polygon were calculated as the shape features serving as the feature vectors to distinguish each polygon again based on the support vector machine method. In this way, the semi-automatic extraction of mural paint loss was realized. A case study was conducted to evaluate the performance of our proposed method, using the hyperspectral data of the Qutan Temple mural. The results show that our proposed method is capable of enhancing the extraction accuracy of mural paint loss and supporting the investigation of the present situation of murals. Related Articles | Metrics

Select

Research and implementation of instance segmentation and edge optimization algorithm LIANG Zheng-xing , WANG Xian-bing , HE Tao , WU Zhong-ding , ZHANG Jia 2020, 41(6): 939-936.  DOI: 10.11996/JG.j.2095-302X.2020060939 Abstract ( 155 )   PDF (2414KB) ( 121 )   Abstract: In recent years, the instance segmentation technology has received more attention. Although the Mask R-CNN instance segmentation method is important in the field of instance segmentation, the resultant edge of each instance cannot entirely match the real edge. In order to solve this problem, a method was proposed that combined the result of the salient object extraction with that of the mask R-CNN instance segmentation, so as to produce a better edge of instance segmentation. First, the image was recognized by Mask R-CNN, with the segmentation result obtained. Then PoolNet was utilized to process the detected image, resulting in the salient object information in the image. At last, the edge of the mask image was optimized by the result of PoolNet, attaining a better result of the edge segmentation. After testing, this method can yield better segmentation results than Mask R-CNN for most of images with salient targets in some important indexes. Related Articles | Metrics

Select

A blind separation algorithm with low complexity for fMRI brain activation  CHEN An-ying, WU Hai-feng, LI Dong 2020, 41(6): 947-953.  DOI: 10.11996/JG.j.2095-302X.2020060947 Abstract ( 46 )   PDF (1251KB) ( 146 )   Abstract: Functional magnetic resonance imaging (FMRI) is a medical imaging technology widely employed in brain region positioning for its non-invasiveness and high spatiotemporal resolution. However, the traditional FMRI signal separation algorithm was too complex and time-consuming to effectively apply the FMRI technology to brain function research. Aiming at the computational complexity of traditional FMRI brain separation algorithms, a blind separation algorithm was proposed based on the second-order Hadamard transform. This algorithm first calculated the correlation function of the blood oxygen level dependent (BOLD) signal in the fMRI data, and then performed eigenvalue decomposition to obtain the unmixing matrix, thereby realizing the activation of brain regions. Given the composition of the Hadamard being only 1 or 1, the complexity can be reduced for the BOLD signal correlation matrix calculation. The simulation results show that compared with the independent component analysis (ICA) of high-order statistics and the Fourier transform blind separation algorithm of second-order statistics, the calculation time of this algorithm was only 25% and 50% of theirs, respectively, while the positioning error was close. Related Articles | Metrics

Select

Large subspace number subspace segmentation via fast convex infinity norm minimization  TANG Ke-wei, MU Meng-jiao, LI Jin-hong, ZHANG Jie, JIANG Wei, PENG Xing-xuan 2020, 41(6): 954-961.  DOI: 10.11996/JG.j.2095-302X.2020060954 Abstract ( 55 )   PDF (479KB) ( 53 )   Abstract: Subspace segmentation is one of the fundamentals in computer vision and machine learning. Given the large number of categories in practical problems concerning the data set, it is of great significance to address the issue of large subspace number subspace segmentation. Although spectral clustering-based methods received more attention in the field of subspace segmentation, the subspace number in the past experiments was usually less than 10. The infinity norm minimization was a recently proposed method specially for large subspace number subspace segmentation. It could effectively address this problem by reducing the difference of the representation matrix, but there remained some limitations. For example, the computation speed was not fast enough, and there was no theoretical guarantee for the independent subspaces. Therefore, a method named fast convex infinity norm minimization was proposed. This method can not only reduce the difference of the representation matrix, but also provide the theoretical guarantee for the independent subspace and enhance the computation speed, which has been testified by a large number of experiments. Related Articles | Metrics

Select

IV LKWA: an information visual analysis tool with advanced L-K optical flow based WebAR  PEI Yun-qiang , WU Ya-dong , WANG Fu-pan , ZHANG Xiao-rong , JIANG Hong-yu , XU Shi-jian , TANG Wen-sheng 2020, 41(6): 962-969.  DOI: 10.11996/JG.j.2095-302X.2020060962 Abstract ( 75 )   PDF (1931KB) ( 65 )   Abstract: For the mobile augmented reality (MAR) technology combined with information visualization, there remains excessive computing pressure on devices in target tracking. If the MAR application adopts the feature tracking solution for 2D images to track 3D objects, the multi-angle feature points in 3D objects obtained from extensive calculations will undoubtedly increase computing pressure in tracking process. As a consequence, excessive computing pressure will be incurred on devices, leading to unstable phenomena in the scene, such as jitter, latency, and movement of 3D AR model lagging behind the target during the target tracking. To resolve these problems, a WebAR (web-based AR) solution was proposed based on the advanced L-K method (Lucas Kanade method, an optical flow algorithm), which transformed the feature point tracking problem into an optical flow estimation problem and an optimized 3D interaction strategy for information visualization. The experimental results could verify that the proposed method can effectively enhance the computing efficiency and stability of target tracking in MAR and enrich the presentation and interaction in information visualization with WebAR. Related Articles | Metrics

Select

Simulation of free interface motion of two-phase flow based on modified pressure and surface tension calculation  ZHU Xiao-lin, ZHOU Yun-ruo, HE Hong-hong 2020, 41(6): 970-979.  DOI: 10.11996/JG.j.2095-302X.2020060970 Abstract ( 77 )   PDF (2168KB) ( 58 )   Abstract: In the simulation of multiphase flow using the smoothed particle hydrodynamics (SPH) method, the density of the multiphase flow at the interface was discontinuous, the pressure calculation of the particle interface was erroneous, and problems occurred, such as pressure oscillation at the interface and interface fracture. In response to the above problems, a new pressure gradient approximation formula and an improved artificial repulsion formula at the interface were proposed to make the interface clearer and smoother without penetration, thus producing better simulation results. In addition, by giving a color function calculation formula based on density weights, the formula was improved for the calculation of surface tension at the interface of multiphase flow with high density ratio, leading to the smoother density transition of multiphase flow interface and better simulation effect. Finally, through three simulation experiments, such as dam break, Rayleigh-Taylor interface instability, and non-Boussinesq locked exchange problem, the particle distribution map of the interface and the distance of the interface front at different times were obtained. The results can verify the rationality of the new pressure gradient approximation formula and the artificial repulsion formula at the interface. Through the simulation experiment of droplet formation in the air, the particle change diagram formed by the circular droplet was obtained, and the result can corroborate the effectiveness of the improved surface tension calculation method in this paper Related Articles | Metrics

Select

A semi-regular mesh simplification algorithm based on inverse Loop subdivision  LUAN Wan-na, LIU Cheng-ming 2020, 41(6): 980-986.  DOI: 10.11996/JG.j.2095-302X.2020060980 Abstract ( 84 )   PDF (980KB) ( 75 )   Abstract: 3D mesh simplification is an operation to minimize the number of vertices and faces in the refined 3D model while preserving the geometric information of the target object. It plays a significant role in improving the access and network transmission speed of the 3D mesh data, and the efficiency of editing and rendering. To address the problem of most mesh simplification algorithms neglecting the mesh topology and visual quality during simplification, a semi-regular mesh simplification algorithm was proposed based on the inverse Loop subdivision. The algorithm first detected the feature points according to the neighborhood centroid offset. Then a seed triangle was randomly selected to obtain the regular region by edge extension, and the inverse Loop subdivision was performed to simplify the mesh. Finally, the simplified model was gained by edge splicing in the way of inwards segmentation. Regarding the open testing data, comparisons were made between the algorithm and the classical ones. The experimental results show that the proposed algorithm can preserve the features effectively and keep the regular topology structure as much as possible during simplification, and that it is superior to the edge collapse and clustering algorithm in visual quality. Related Articles | Metrics

Select

Algorithm for calculating the shortest distance between the inner surfaces of the pocket of an aircraft structure  HU Bao-ying , ZHANG Tian-yang , ZHENG Guo-lei , ZHOU Min 2020, 41(6): 987-992.  DOI: 10.11996/JG.j.2095-302X.2020060987 Abstract ( 51 )   PDF (619KB) ( 58 )   Abstract: In the development process of the automatic programming system for computer numerical control (CNC) machining of aircraft structural parts, the process defects of the part feature, especially the shortest distance between the inner surfaces of the pockets being too small, was likely to complicate the processing of parts, incur high cost, or even render the processing impossible. At present, this model problem was mainly inspected manually, which was inefficient and error-prone. For the pockets of aircraft structural parts with complex shapes and such complex features as open and closed angles and indentations, it was difficult to determine the shortest distance between side walls. Based on the extraction of feature geometric parameters, the calculation of the shortest distance between side walls was converted to the computation of the shortest distance between lines. The validity rules of the distance calculation between lines were presented to simplify the calculation according to geometric features of the pocket’s characteristics, and the shortest distance between inner wall surfaces of pocket features was obtained. Finally, an aircraft structure model is taken as an example to verify the feasibility and effectiveness of the algorithm. Related Articles | Metrics

Select

Design of aided mechanism configuration for human body turning-over based on motion trajectory fitting SU Peng, LU Da , LUN Qing-long , LI Jian , XU Xiao-zhong , FAN Yu-bo, 2020, 41(6): 993-1000.  DOI: 10.11996/JG.j.2095-302X.2020060993 Abstract ( 63 )   PDF (3928KB) ( 245 )   Abstract: Turning over from supine position is one of effective measures to prevent pressure ulcers for long-term bedridden patients, and it is significant to explore mechanisms of human movement and to personate the design of aided human turnover mechanism. The human shoulders and buttocks are the key force application positions for the aided human turnover. Based on the human anatomy, the spatial motion of the shoulder and buttock bones were analyzed, and the motion marker model was established. In addition, the motion capture experiment of the human supine turnover position was undertaken to obtain figures of the motion trajectory and the feature information of the linkage model of shoulders and buttocks. Through the analysis on the changing progress of the length and angle of the linkage mechanism, the graph geometric analysis and simplification of the linkage mechanism were carried out. Based on the motion trajectory graph fitting, the configuration design of the aided mechanism was completed, with the kinematics simulation implemented. The slight error of simulation and experiment graph fitting proves the rationality of the proposed anthropomorphic aided mechanism method. The results can provide a theoretical basis for the design of the man-machine collaborative rehabilitation mechanism. Related Articles | Metrics

Select

Research on establishment of city geographic information platform based on open source technology  LENG Shuo , LI Sun-wei , HU Zhen-zhong, 2020, 41(6): 1001-1011.  DOI: 10.11996/JG.j.2095-302X.2020061001 Abstract ( 87 )   PDF (2502KB) ( 93 )   Abstract: To solve the problems in the establishment of city geographic information platform including difficulty in data acquisition and high development cost, this paper introduced open source technologies, and proposed a platform establishment method based on open source data, open source tools and open source framework. The proposed platform consists of two parts, including the data module and the visualization module. In order to construct the data module, the open source geographic project open street map (OSM) was selected to obtain building and division data, and a building merging algorithm was designed to further improve data quality. The building data was then organized hierarchically based on administrative divisions and provided to specific applications through the data interface constructed by ASP. NET. The visualization module was implemented based on the open source geographic information system (GIS) project Cesium, and complete user interfaces were designed. Taking Beijing as an example, we carried out the development and testing of the prototype system, and confirmed the feasibility of the proposed technical route. The advantages of open source technologies were fully considered in this paper to reduce platform development costs and promote the research and application of city geographic information platform.  Related Articles | Metrics

Select

Research on external vehicle interaction design based on AAM model   ZHANG Bi-han, YOU Fang 2020, 41(6): 1012-1017.  DOI: 10.11996/JG.j.2095-302X.2020061012 Abstract ( 97 )   PDF (1173KB) ( 91 )   Abstract: With the continuous advancement of science and technology, new technologies have been constantly ushered into people’s lives, and the autonomous driving technology has also been developing rapidly. However, the technological development alone cannot increase people’s acceptance. In order to improve the acceptance of unmanned driving technology, the automation acceptance model was studied. Explorations were conducted on how to combine the acceptance model with interaction design, and analyses were made regarding the user experience of external vehicle interaction based on the acceptance model. Through the analysis of the use scenarios of external vehicle interaction in semi-closed scenes and the user journey map during pedestrians’ interaction with automation vehicles, a pedestrian behavior characteristic model was summarized. In addition, on the basis of related research on the automation acceptance model, the effectiveness and acceptance of external vehicle interaction were improved from the perspective of trust and compatibility. Based on the AAM model, a strategyfitting the pedestrian behavior characteristics and cognition was proposed for human-machine external vehicle interaction, thereby enhancing communication efficiency and acceptance. Related Articles | Metrics

Select

Research on computer aided recognition method of bronze decoration feature contour   ZONG Li-cheng , WANG Na-na 2020, 41(6): 1018-1023.  DOI: 10.11996/JG.j.2095-302X.2020061018 Abstract ( 98 )   PDF (805KB) ( 93 )   Abstract: Centering on the key algorithms and techniques for identifying the contour lines of bronze ornamental features, analyses were made on the present situation of the digital design of current decorations. Based on the computer-aided technology and triangular grid basic theory, using the image clustering and threshold method, it was proposed that the key technology of characteristic contour recognition comprised contour corner point, characteristic curve segment, segmented contour matching, and spatial curve matching. The recognition algorithm and process for the target feature of the decoration were constructed, and the maximum curvature was employed to identify contour corners. In addition, the characteristic curve segment was utilized to express and project the contour curve, and the curve chord length method was adopted to match and splice the characteristic curve. A method was established for identifying bronze decorative patterns, and based on computer-aided technology, the complete process was realized from pattern recognition to spatial curve matching and splicing. The bronze feature recognition method based on the computer-aided design technology possessed the characteristics of digitization, high-level intelligence and efficiency. The experimental results indicate that there is a promising prospect for the application of this method in target object pattern recognition and extraction. Related Articles | Metrics

Select

Extension reconstruction method of design factors for similar images in cultural and creative graphics   WANG Wei-wei , SONG Zhen-zhen , LI Pei 2020, 41(6): 1024-1030.  DOI: 10.11996/JG.j.2095-302X.2020061024 Abstract ( 121 )   PDF (3910KB) ( 201 )   Abstract: At present, the prosperity and development of China’s cultural and creative industries have greatly boosted the inheritance and promotion of traditional Chinese culture. In order to realize the style-oriented design scheme of cultural and creative products according to users’ expectation, the reliance on the subjective design consciousness of designers should be reduced, and explorations should be made on the approach to transforming traditional cultural design factors into modern innovative design factors. Firstly, the Illustrator software was employed to extract samples of traditional cultural elements as design factors, image adjectives were collected, and the image value of the unit design factors in the sample library was evaluated using the SD method, experimental method, and statistical theory, thus determining the similar images and the main and auxiliary images of the unit design factor. Secondly, through the relationship between image and design factors, a search database based on Relational database management system (RDBMS) was established, and the similar-image design factors expected by users were indexed and analyzed using the extension theory and diamond thinking model to achieve the extension transformation reconstruction design. Finally, the feasibility of the method and process in this research was verified by the reconstruction process and design application of the cultural elements of the Forbidden City. Related Articles | Metrics

Select

Research on project-driven digital image processing teaching with the OBE concept  CUI Wen-chao, ZOU Jun-jie, WANG Fang-yi, TANG Ting-long, XIA Ping 2020, 41(6): 1031-1038.  DOI: 10.11996/JG.j.2095-302X.2020061031 Abstract ( 173 )   PDF (756KB) ( 117 )   Abstract: The traditional teaching of digital image processing emphasized the explanation of classic methods, the implementation of each single algorithm, and the training of individual abilities, while neglecting the expansion of new methods in literature, the project development of algorithms, and the cultivation of team spirit. In order to overcome these problems, a new project-driven teaching mode was explored based on the outcomes-based education (OBE) concept certified by the profession of engineering education. Concretely, the teaching goals were determined by the index points of the graduation requirements for electronic information majors. Using the project teaching method of CDIO, the teaching process was divided into four stages: project investigation, project design, project implementation, and project operation feedback. A special subject on image thresholding segmentation was demonstrated as an example to explain in detail the works conducted and abilities/qualities fostered/trained in different stages. As a result, the entire teaching process centered on ability training and quality education. Finally, the assessment and feedback on the quality of teaching and learning were employed to constantly address the shortcomings in the development of students’ skills. Through four years of teaching practice, the project-driven teaching method based on the OBE concept has been proven effective not only in overcoming the disadvantages of the traditional teaching method, but also in enhancing students’ comprehensive ability and quality. Related Articles | Metrics

Select

Application research on the mixed teaching mode in the mechanical graphics courses ZHOU Rong-an, FU Chun-ming 2020, 41(6): 1039-1043.  DOI: 10.11996/JG.j.2095-302X.2020061039 Abstract ( 146 )   PDF (4017KB) ( 133 )   Abstract: To address students’ lack of spatial imagination and learning initiative and the difficulty in teach-student interactions in the traditional teaching of mechanical graphics, this paper mainly drew from the reform experience of the American universities: the “active and collaborative learning” integrated teaching modes. In the teaching process, the real time interactions in class and “online + offline” modes after class were utilized to foster students’ active studying ability. At the same time, the collaborative learning assisted by the creative design projects was employed to train students’ team collaboration and innovation spirits. Take one class from University of South China as the object of this study, the results of the practices indicate that the teaching proposal could enhance the effectiveness and quality of teaching. Related Articles | Metrics