模式识别与人工智能
Home      About Journal      Editorial Board      Instructions      Ethics Statement      Contact Us                   中文
Pattern Recognition and Artificial Intelligence
22 Judgement and Disposal of Academic Misconduct Article
22 Copyright Transfer Agreement
22 Proof of Confidentiality
22 Requirements for Electronic Version
More....
22 Chinese Association of Automation
22 National ResearchCenter for Intelligent Computing System
22 Institute of Intelligent Machines,Chinese Academy of Sciences
More....
 
 
2016 Vol.29 Issue.3, Published 2016-03-31
Papers and Reports    Researches and Applications   
   
Papers and Reports
193 Stereo Matching Based on Adaptive Matching Windows and Multi-feature Fusion
SHI Hua, ZHU Hong
Aiming at the problems of local stereo matching methods, such as difficulties in matching window selection, vague disparity of edges and low accuracy in weak texture regions and slope surface regions, an efficient stereo matching algorithm with adaptive support window based on segmentation in CIELAB color space and multiple features fusion is proposed in this paper. Firstly, the stereo images are segmented in CIELAB color space, the initial support window is calculated according to the homogeneous regions, and the initial support window is updated by estimating the occlusion region. And then the initial disparity map in the updated support region is achieved by the linear weighted multi-feature fusion matching method with adaptive weights. Finally, the mismatch is checked by consistency of right disparity and left disparity, and then the ultimate dense disparity map is obtained through disparity optimization by mean filtering and disparity refinement. Experimental results show that the proposed algorithm is effective with high matching precision, especially for weak texture and slope surface regions.
2016 Vol. 29 (3): 193-203 [Abstract] ( 795 ) [HTML 1KB] [ PDF 1373KB] ( 826 )
203 Assembly Ontology for Assembly Sequence Planning
MENG Yu, GU Tianlong, CHANG Liang, LI Fengying

The ontology technology is introduced into the assembly modeling to improve the automation degree of the assembly sequence planning, realize the information sharing, reusing and transferring in related activities of the assembly, and implement the seamless integration and mutual operation between the assembly sequence planning system and other heterogeneous systems. Aiming at the modeling demand of the assembly sequence planning, the assembly sequence planning-oriented and assembly object-based assembly ontology named web ontology language for assembly sequence planning (OWL-ASP) is constructed. The OWL-ASP consists of Properties ontology, AssemblyInfo ontology and AssemblySeq ontology and theydescribe the assembly properties, assembly information and assembly sequence of the assembly object, respectively. Based on OWL-ASP, the rules for assembly are depicted with the semantic web rule language(SWRL) to form the unified hierarchical system for assembly knowledge representation and realize the strict depiction and automated reasoning of the assembly knowledge. Effectiveness of the proposed ontologies and rules for assembly is illustrated through a practical example of ball valve assembly.
2016 Vol. 29 (3): 203-215 [Abstract] ( 489 ) [HTML 1KB] [ PDF 823KB] ( 681 )
216 Two-Stage Score Normalization Method for Spoken Term Detection
LI Peng, QU Dan
Score normalization is an essential part for a spoken term detection (STD) system. In this paper, a two-stage score normalization method is proposed. Firstly, two features, rank-p and relative-to-max, are introduced into a discriminative score normalization method to get more discriminative confidence scores between correct and wrong candidate words, and thus the keyword verification is more accurate. Secondly, a term-weighted value evaluation metric based normalization method is applied to further optimize the performance of STD. Experimental results show that the proposed method takes advantages of both discrimination and metric-based score normalization methods, and it obtains better performance than the best single score normalization method does.
2016 Vol. 29 (3): 216-222 [Abstract] ( 417 ) [HTML 1KB] [ PDF 369KB] ( 650 )
223 (α,β)-Quasi-Lock-Semantic Resolution Method of Intuitionistic Fuzzy Logic
ZOU Li, LIU Di, TAN Xuewei, ZHENG Hongliang
To improve the (α,β)-resolution efficiency of intuitionistic fuzzy propositional logic, quasi lock semantic resolution is applied to (α,β)-resolution. The (α,β)-quasi lock semantic resolution method is introduced into intuitionistic fuzzy logic system, and its soundness and completeness are proved. The concepts of (α,β)-quasi lock semantic resolution and (α,β)-quasi lock semantic resolution reasoning of intuitionistic fuzzy propositional logic are proposed. The formula of (α,β)-quasi lock semantic resolution and merger rule of generalized lock clauses are discussed. Finally, the steps of automated reasoning algorithm based on (α,β)-quasi lock semantic resolution for intuitionistic fuzzy propositional logic are presented and an example is given to illustrate the effectiveness of the proposed method.
2016 Vol. 29 (3): 223-228 [Abstract] ( 481 ) [HTML 1KB] [ PDF 358KB] ( 462 )
229 Real-Time Multi-scale Object Tracking Based on Cluster Similarity
LI Kang, HE Fazhi, CHEN Xiao, PAN Yiteng, YU Haiping
To solve the problems of real-time object tracking and scale changing of the object in object tracking, a real-time object tracking algorithm is proposed based on cluster similarity measurement (MSCSM) in particle filtering framework. The improved average haar-like features are utilized to represent the proposed appearance model. Firstly, the target cluster and the background cluster are cropped in their sample radii. Secondly, a similarity measurement between a particle and a cluster is defined. The score of each particle is calculated according to its similarity with clusters while a new frame coming. Finally, the particle with the maximum score is selected as the new target location in the current frame. At the end of tracking for each frame, the statistical characteristics of clusters are updated and the particles are resampled to avoid degeneration.The proposed algorithm shows superiority in comparison with the state-of-the-art algorithms.
2016 Vol. 29 (3): 229-239 [Abstract] ( 658 ) [HTML 1KB] [ PDF 3546KB] ( 671 )
Researches and Applications
240 Multi-label Feature Selection Algorithm Based on Local Subspace
LIU Jinghua, LIN Menglei, WANG Chenxi, LIN Yaojin
In the existing multi-label feature selection algorithms, the features with stronger relevance to label set are usually selected according to some related criteria. However, this strategy may not be the optimal option. As some features may be the key features for a few labels, but they are weakly related to the whole label set. Based on this assumption, a multi-label feature selection algorithm based on local subspace is proposed. Firstly, the mutual information between feature and label set is employed to measure the importance degree of each feature, and original feature sequences are ranked by their importance degree from high to low to obtain a new feature space. Then, the new feature space is partitioned into several subspaces, and the less redundant features are selected in each subspace by setting a sampling ratio. Finally, the final feature subset is obtained by merging all feature subsets in different subspaces. Experiment is conducted on six datasets and four evaluation criteria are used to measure the effectiveness. Experimental results show that the proposed algorithm is superior to the state-of-the-art multi-label feature selection algorithms.
2016 Vol. 29 (3): 240-251 [Abstract] ( 722 ) [HTML 1KB] [ PDF 3526KB] ( 696 )
252 Granular Computing Based Hesitant Fuzzy Multi-criteria Decision Making
WANG Baoli, LIANG Jiye, HU Yunhong
A multi-criteria decision making method is proposed based on granular computing in hesitant fuzzy setting. Firstly, the possibility degree is defined for comparing hesitant fuzzy elements in a hesitant fuzzy set and an additive consistent fuzzy preference matrix is constructed based on the defined possibility degrees. Secondly, the criteria weights are determined by the preorder entropies and the similarity degree of preorder granular structures. Thirdly, an ordered vector is obtained by integrating the corresponding fuzzy preference matrices with the criteria weights. The criteria weights are computed from the evaluation information content and the relations between orders of the individual criteria and the whole evaluation system. The final ranking is obtained by weighting the ranking order under each criterion. Finally, a case study is utilized to verify the effectiveness and feasibility of the proposed method.
2016 Vol. 29 (3): 252-262 [Abstract] ( 600 ) [HTML 1KB] [ PDF 400KB] ( 470 )
263 Leaf Shape Description and Classification with LDA Topic Mode
YE Xulun, ZHAO Jieyu, CHEN Nenglun
Since the conventional shape descriptors focus on shape reexpression and the stochastic character of leaf shape is neglected, an effective plant leaf shape descriptor is proposed based on latent Dirichlet allocation (LDA) model. A corresponding leaf shape recognition framework is also constructed. Firstly, the plant leaf shape is transformed and represented as a multi-scale bag-of-words model, and thus the space interaction relationship is introduced into the leaf shape generative model. Furthermore, a leaf shape generative model is established via LDA model, and then the leaf shape descriptor is designed by the extracted shape distribution parameters in the LDA model. Finally, k-nearest neighbor (KNN) method is applied to the plant leaf shape classification. Experimental results demonstrate that the leaf shape descriptor combined with LDA model effectively improves the shape classification accuracy, especially for the plants of different classes but with a roughly similar shape of leaf. The proposed method obtains a higher classification accuracy compared with the conventional shape descriptor.
2016 Vol. 29 (3): 263-271 [Abstract] ( 591 ) [HTML 1KB] [ PDF 801KB] ( 728 )
272 Optimal Scale Selection in Multi-scale Contexts Based on Granular Scale Rules
HAO Chen, FAN Min, LI Jinhai, YIN Yunqiang, WANG Dujuan
Firstly, several kinds of multi-scale contexts are defined firstly, the notion of a scale rule is put forward, and some properties of scale rules are discussed. Secondly, decision scale is introduced into multi-scale contexts to form multi-scale decision contexts, and the redundancy between scale rules is also investigated. Moreover, granular scale rules are employed to define the consistency of multi-scale decision context, and an optimal scale selection method is presented grounded on the consistency guarantee of the multi-scale decision context. Finally, numerical experiments show the effectiveness of the proposed method.
2016 Vol. 29 (3): 272-280 [Abstract] ( 592 ) [HTML 1KB] [ PDF 391KB] ( 633 )
281 Collaborative Filtering Recommendation Algorithm Incorporating Social Network Information
GUO Lanjie, LIANG Jiye, ZHAO Xingwang
To solve the problems of high data sparsity and limited recommendation precision of collaborative filtering recommendation algorithms, a collaborative filtering algorithm incorporating social network information is proposed under the framework of item-based collaborative filtering recommendation. In item similarity calculation period and user rating prediction period, social network information is utilized to fill missing values in rating matrix selectively and thus the existing rating information is utilized as much as possible. Finally, experiment is conducted on Epinions dataset. Results show that the proposed algorithm alleviates the data sparsity problem and outperforms other collaborative filtering algorithms on rating error and precision.
2016 Vol. 29 (3): 281-288 [Abstract] ( 646 ) [HTML 1KB] [ PDF 472KB] ( 981 )
模式识别与人工智能
 

Supervised by
China Association for Science and Technology
Sponsored by
Chinese Association of Automation
NationalResearchCenter for Intelligent Computing System
Institute of Intelligent Machines, Chinese Academy of Sciences
Published by
Science Press
 
Copyright © 2010 Editorial Office of Pattern Recognition and Artificial Intelligence
Address: No.350 Shushanhu Road, Hefei, Anhui Province, P.R. China Tel: 0551-65591176 Fax:0551-65591176 Email: bjb@iim.ac.cn
Supported by Beijing Magtech 　Email:support@magtech.com.cn