Fast template tracking in video sequences

Tracking of poorly defined, rotating and/or distorted objects in a video sequence poses significant problems especially in medical diagnostics including ultrasound (sonographic) video used for examination and diagnosis of internal movement of tissue or muscle and nerve action. Cross-correlation techniques have been successful in retrieving dynamic information directly from ultrasound video data. We outline a fast implementation of tracking based on normalized cross-correlation using an adaptive template and present results from our application, developed in MATLAB?, which successfully tracks arbitrarily selected objects in deformed or severely compromised images. Common ultrasound image evaluation is qualitative but there is need to retrieve quantitative dynamic information such as the trajectory or velocity of selected areas. Our approach uses normalized two-dimensional cross-correlation to find the position of an initially selected template enclosing the feature of interest and map its trajectory frame-by-frame to produce displacement and velocity plots. We illustrate operation of the application using routine ultrasound data and demonstrate its performance using test video of objects rotating full circle and rolling down a ramp. We analyse errors associated with sampling to compare performance of our implementation with a more rigorous but tedious and computationally expensive correlation of a resampled, rotated, and shifted template.     

机场免费WiFi安全吗？

通过一个开放的WiFi网络访问互联网是危险的行为,因为你并不知道谁是服务提供商,并且你将会连接到何处。在机场里,你可以通过机场的免费WiFi来收发邮件或者更新你的Facebook,但是如果这么操作,你的这些终端将有可能无法保障安全。     

Review of a pivotal Human Factors article: "Humans and automation: use, misuse, disuse, abuse"

OBJECTIVE: This paper considers the influence of "Humans and Automation: Use, Misuse, Disuse, Abuse" and examines how it relates to the evolving issue of human-automation interaction. BACKGROUND: Automation presents important practical challenges that can dramatically affect satisfaction, performance, and safety; philosophical challenges also arise as automation changes the nature of work and human cognition. METHOD: Papers cited by and citing "Humans and Automation" were reviewed to identify enduring and emerging themes in human-automation research. RESULTS: "Humans and Automation" emerges as an important node in the network of automation-related papers, citing many and being cited by many recent influential automation-related papers. In their article, Parasuraman and Riley (1997) integrated previous research and identified differing expectations across designers, managers, and operators regarding the need to support operators as a source of automation problems. They also foresaw and inspired research that addresses problems of overreliance and underreliance on automation. CONCLUSION: This pivotal article and associated research show that even though automation seems to relieve people of tasks, automation requires more, not less, attention to training, interface design, and interaction design. The original article also alludes to the emergence of vicious cycles and dysfunctional meta-control. These problems reflect the coevolution of automation and humans, in which both adapt to the responses of the other. APPLICATION: Understanding this coevolution has important philosophical implications for the nature of human cognition and practical implications for satisfaction, performance, and safety.     

Continuous State Dynamic Programming via Nonexpansive Approximation

This paper studies fitted value iteration for continuous state numerical dynamic programming using nonexpansive function approximators. A number of approximation schemes are discussed. The main contribution is to provide error bounds for approximate optimal policies generated by the value iteration algorithm.      

Extracting Takagi-Sugeno Fuzzy Rules with Interpretable Submodels via Regularization of Linguistic Modifiers

In this paper, a method for constructing Takagi-Sugeno (TS) fuzzy system from data is proposed with the objective of preserving TS submodel comprehensibility, in which linguistic modifiers are suggested to characterize the fuzzy sets. A good property held by the proposed linguistic modifiers is that they can broaden the cores of fuzzy sets while contracting the overlaps of adjoining membership functions (MFs) during identification of fuzzy systems from data. As a result, the TS submodels identified tend to dominate the system behaviors by automatically matching the global model (GM) in corresponding subareas, which leads to good TS model interpretability while producing distinguishable input space partitioning. However, the GM accuracy and model interpretability are two conflicting modeling objectives, improving interpretability of fuzzy models generally degrades the GM performance of fuzzy models, and vice versa. Hence, one challenging problem is how to construct a TS fuzzy model with not only good global performance but also good submodel interpretability. In order to achieve a good tradeoff between GM performance and submodel interpretability, a regularization learning algorithm is presented in which the GM objective function is combined with a local model objective function defined in terms of an extended index of fuzziness of identified MFs. Moreover, a parsimonious rule base is obtained by adopting a QR decomposition method to select the important fuzzy rules and reduce the redundant ones. Experimental studies have shown that the TS models identified by the suggested method possess good submodel interpretability and satisfactory GM performance with parsimonious rule bases.     

Copyright Infringement and Protection in the Internet Age

In this age of increasingly complex, Internet-based technologies, it's very difficult to protect original works of authorship. This article examines the problem of copyright infringement and the copyright law's desire to balance the interests of creators with the interests of the public. The author also offers some insight in how to reduce the problem of copyright infringement.     

Error compensation in machine tools: a neural network approach

For a non-idealized machine tool, each point in the workspace is associated with a tool point positioning error vector. If this error map can be determined, then it is possible to substantially improve the positioning performance of the machine by introducing suitable compensation into the control loop. This paper explores the possibility of using an artifical neural network (ANN) to compute this mapping. The training set for the ANN is obtained by mounting a physical artifact whose dimensions are precisely known in the machine's workspace. The machine, equipped with a touch trigger probe, measures the positions of features on the artifact. The difference between the machine reading and the known dimension is the machine error at that point in the workspace. Using standard modeling techniques, the kinematic error model for a CNC turning center was developed. This model was parameterized by measurement of the parametric error functions using a laser interferometer, electronic levels and a precision square. The kinematic model was then used to simulate the artifact-measuring process and develop the ANN training set. The effect of changing artifact geometry was explored and a machining operation was simulated using the ANN output to provide compensation. The results show that the ANN is capable of learning the error map of a real machine, and that ANN-based compensation can significantly reduce part-dimensional errors.     

Exploring the Semi-Supervised Video Object Segmentation Problem from a Cyclic Perspective

International Journal of Computer Vision - Modern video object segmentation (VOS) algorithms have achieved remarkably high performance in a sequential processing order, while most of currently...     

Dynamics of Plug Formation in a Circular Cylinder Under Low Bond Number Conditions: Experiment and Simulation

The goal of the current study is to investigate the dynamics of two phase interface under a low Bond number condition. Silicone oil is injected into a cylinder under a Bond number of about 0.47 via a side tube forming a T-junction with the former. The time evolution of the interface of silicon oil in a cylinder is captured using a high speed camera. The volume at which the plug is formed is then determined using an image processing tool to analyze the captured images. A numerical simulation is carried out where fluid is injected into a cylinder, under a less than unity Bond number condition, via a side tube. Numerical and experimental results are then compared.