This special issue of IJCM offers a broad platform to present innovative approaches and technologies in visual information analysis and retrieval. It is obvious that mathematical modelling, applied mathematics methods, and computational techniques of visual information analysis and retrieval play an important role in the recent development of this field and have been changing our everyday life dramatically over the last decade. In the sequel, we introduce the origin of this special issue and then give an overview of the papers included.
The papers emanate from the International Conference on Image and Video Retrieval (CIVR 2010), which was held on 3–7 July 2010 in Xi'an, China. The aim was to provide a high-level international forum for scientists and researchers to discuss the state of image and video retrieval and bring together scientists from the international community in this field in order to present their latest research work. The conference also provided a good platform for researchers to share their new ideas, experience and progress and to explore novel directions and problems in this fascinating area. There were more than 100 participants from 18 countries and regions at this conference. The CIVR 2010 attracted 153 paper submissions from 20 countries. From these 153 papers, 53 were accepted, where 16 were oral and 37 were poster. The proceedings were published by ACM. Thereafter, out of these 53 papers, 12 of high quality were selected for this special issue. These papers were then substantially extended and these went through another round of reviewing for rigorous peer review. These papers cover comprehensive aspects of visual information analysis and retrieval.
This special issue starts with three papers on video content analysis. The first one focuses on the classification of video events based on evolutionary optimization Citation9. The second one concentrates on near-duplicate video detection, which has become an important task in many multimedia applications Citation6. Paisitkriangkrai et al. propose a practical solution for the real-time near-duplicate video detection in large-scale video databases based on a more discriminative signature of video clips. The third one involves video copy detection by considering both visual and audio information simultaneously Citation7. A bag-of-visual concept words model and a bag-of-audio words model for video copy detection are introduced from visual and audio cues. A coherency vocabulary combined with soft-weighted strategy realizes a fast and accurate indexing, and the late fusion is adopted to obtain the final audio–visual result based on visual-only and audio-only results.
The next three papers are related to image classification and indexing. The first paper, ‘Results selection diversity for web image retrieval’ Citation5, describes a re-ranking method, called dual-rank, to improve web image retrieval by clustering and re-ordering the images retrieved from an image search engine. The second paper is ‘Dayside corona aurora classification based on X-grey level aura matrices and feature selection’ Citation11, in which corona aurora is the main form of aurora at magnetic noon generated by the dynamics process of the interaction of the sun and the earth's magnetosphere. This paper proposes a novel aurora texture description method based on X-grey level aura matrices and feature selection. The third paper proposes a hybrid relevant-diverse image re-ranking approach, which is a cluster-based re-ranking method Citation1.
There are two papers dedicated to human action representation and recognition, including human action representation using pyramid correlogram of oriented gradients on motion history images Citation8 and action recognition using graph embedding and the co-occurrence matrix descriptor Citation12.
Another two papers deal with image processing, that is, image fusion and watermarking. In the paper ‘An application of compressive sensing for image fusion’ Citation10, Wan and Qin propose a study of three sampling patterns and investigate their performance on compressive-sensing-based image reconstruction. Then, a new image fusion algorithm is proposed in the compressive domain by using an improved sampling pattern. The paper ‘Robust curvelet-domain image watermarking based on feature matching’ Citation4 presents a copyright protection method with digital watermarking in one area of multiscale geometric analysis, that is, the curvelet domain.
The final two papers are concerned with applications of the human visual system, that is, object detection and visual attention. The paper ‘Saliency based on cortex-like mechanisms’ Citation3 presents the saliency criteria to measure the perspective fields. A saliency criterion is obtained from two pathways, Shannon's information entropy and independent component analysis, to model a visual attention system, which can locate the most informative spots in complex environments. The paper ‘Mining concise and distinctive affine-stable features for object detection in large corpus’ Citation2 gives a novel algorithm for mining concise and distinctive invariant features, called affine-stable characteristics. Two new notions, global stability and local stability, are introduced to calculate the robustness of each feature from two mutually complementary aspects. Furthermore, to make these stable characteristics more distinctive, spatial information taken from several representative scales is encoded in a concise way.
It is expected that all these papers will be beneficial to readers interested in the field of visual information analysis and retrieval. As Guest Editors, and on behalf of all the authors of this special issue, we sincerely thank Professor George Loizou for his constant support during the process of preparation and publication of this special issue. We also express our appreciation to the authors for their continuous effort and enduring patience as this special issue came together and the reviewers who have given helpful comments and suggestions on the selected submissions. We are grateful to the Natural Science Foundation of China (NSFC) and all co-sponsors for their support of CIVR 2010.
References
- Cui , C. , Ma , J. , Zhang , L. , Li , P. and Ren , Z. 2011 . A hybrid relevant-diverse approach for image re-ranking with multiple features . Int. J. Comput. Math. , 88 ( 18 ) : 3864 – 3881 .
- Gao , K. , Zhang , Y. , Zhang , W. and Lin , S. 2011 . Mining concise and distinctive affine-stable features for object detection in large corpus . Int. J. Comput. Math. , 88 ( 18 ) : 3953 – 3962 .
- Han , B. , Gao , X. , Tcheang , L. and Walsh , V. 2011 . Saliency based on cortex-like mechanisms . Int. J. Comput. Math. , 88 ( 18 ) : 3942 – 3952 .
- Ji , F. , Huang , D. , Deng , C. , Zhang , Y. and Miao , W. 2011 . Robust curvelet-domain image watermarking based on feature matching . Int. J. Comput. Math. , 88 ( 18 ) : 3931 – 3941 .
- Li , P. , Ma , J. and Zhang , L. 2011 . Results selection diversity for web image retrieval . Int. J. Comput. Math. , 88 ( 18 ) : 3834 – 3851 .
- Liu , Y. , Xu , C. and Lu , H. 2011 . Audio-visual large-scale video copy detection . Int. J. Comput. Math. , 88 ( 18 ) : 3803 – 3816 .
- Paisitkriangkrai , S. , Mei , T. , Zhang , J. and Hua , X.-S. 2011 . Clip-based hierarchical representation for near-duplicate video detection . Int. J. Comput. Math. , 88 ( 18 ) : 3817 – 3833 .
- Shao , L. , Zhen , X. , Liu , Y. and Ji , L. 2011 . Human action representation using pyramid correlogram of oriented gradients on motion history images . Int. J. Comput. Math. , 88 ( 18 ) : 3882 – 3895 .
- Tahayna , B. , Belkhatir , M. , Alhashmi , S. M. and O'Daniel , T. 2011 . Evolutionary optimization of video event classification . Int. J. Comput. Math. , 88 ( 18 ) : 3784 – 3802 .
- Wan , T. and Qin , Z. 2011 . An application of compressive sensing for image fusion . Int. J. Comput. Math. , 88 ( 18 ) : 3915 – 3930 .
- Wang , Y. , Li , J. , Fu , R. and Han , B. 2011 . Dayside corona aurora classification based on X-grey level aura matrices and feature selection . Int. J. Comput. Math. , 88 ( 18 ) : 3852 – 3863 .
- Zheng , F. , Shao , L. , Song , Z. and Chen , X. 2011 . Action recognition using graph embedding and the co-occurrence matrices descriptor . Int. J. Comput. Math. , 88 ( 18 ) : 3896 – 3914 .