Ieee Transactions on Industrial Informatics 1 Guest Editorial Special Section on Intelligent Video Systems and Analytics

I NTELLIGENT video systems and analytics technology have been substantially driven and demanded by recent advanced electronics technology, industrial applications, and the booming market. This Special Section highlights the state-of-the-art and future challenges via the eleven selected papers. The first two papers contribute to video systems architecture and real-time classification/recognition application. Wang et al. develop an application-oriented service share model, enabling various computer vision algorithms integrated into a workable system with high scalability. Kafai and Bhanu demonstrate a robust and complete video-based vehicle classification system based on the rear-side view. Si et al. integrate a set of novel algorithms to effectively enhance the overall performance of iris recognition systems. This Special Section also presents recent efforts on solving classic video problems. Chen et al. improve background subtraction algorithm based on the combination of segmented regions and pixel descriptors to accommodate the conditions of different objects. Tsai et al. present a dissimilarity measure based on the optical-flow technique for surface defect detection. Tan et al. combine smoothen 2D histogram and Gaussian model for automatic skin detection in color images to achieve lower computational cost with improved recognition accuracy. Zhang et al. propose a method extended from fuzzy genetic algorithm to boost the computing efficiency of covariance matching, its performance is significantly superior to its counterparts using exhaustive search. Jin et al. present real-time face detection using cascades of boosted classifiers. The last three papers give priority to the fast-growing topics of 3D gesture recognition and 3D reconstruction. Cao et al. identify key poses by utilizing manifold learning technique to recover the geometric structure of descriptors in a lower dimensional manifold space. Tran and Trivedi present a 3D gesture recognition approach for human computer interactivity based on marker-less upper body pose tracking with 90% average classification accuracy. Nielsen et al. demonstrate an automated technique on thinning of peach blossoms on perpendicular V-architecture trees via a correlation-based stereo mapping. This Special Section is timely and has achieved its aim with profound responses in booming research and technology transfer in intelligent video/camera systems and analytics, it has been successfully completed with cross-disciplinary efforts. We would like to, first of all, thank all the authors who have submitted papers to this Special Section for their contribution and support in intelligent video systems and analytics. Second, we are truly grateful for the efforts and patience of the reviewers whose expertise guarantee the outstanding quality of the …