Detecção de Movimento

Artigos sobre detecção de movimento

<bibtex>@article{Custance, title = {A method for evaluating video motion detection}, author = {Custance, N.D.E. and Devereux, H.}, journal = {Security Technology, 1995. Proceedings. Institute of Electrical and Electronics Engineers 29th Annual 1995 International Carnahan Conference on}, abstract = {Over the last three years there has been an expansion of equipment offered on the market with the capability of providing video motion detection or intelligent scene monitoring. There is currently no easy way to compare this type of equipment due to the many variables in the scenes and the differing algorithms and approaches used in equipment. The Security Equipment Assessment Laboratory has put together a methodology for the testing of video processing equipment. This technique utilises real imagery and provides a basic grounding for comparison. The problems presented by a large library of video sequences on tape are the time taken to run tests, and the relative degradation with time of the video signal from the tape. In particular the degradation in frame stability. The paper proposes formats for attack trials, the recording and playback process control and the logging equipment required. These images are then available for quick evaluation of other video processing systems for particular scenarios. This paper breaks down the key scenarios, and the basis for the selection of video sequences. It also presents the statistical selection process used to identify the more useful image sequences for transfer to optical media. The terms “probability of detection” and “false alarm rate,” can be misleading in the context of video processing systems where different scenarios and attack methods are used. This paper outlines an alternative method for comparing systems based on the content of the scenarios, allowing a prediction of the detection ability of the system and the likely false alarm rate in any situation}, }</bibtex>

<bibtex>@article{He, title = {Evaluation of motion detection techniques for video surveillance}, author = {He, Fangpo and Fettke, M. and Naylor, M. and Sammut, K.}, journal = {Information, Decision and Control, 2002. Final Program and Abstracts}, abstract = {Video motion detection is fundamental in many autonomous video surveillance strategies. However, in outdoor scenes where inconsistent lighting and unimportant, but distracting, background movement is present, it is a challenging problem. Recent research has produced several background modelling techniques, based on image differencing, that exhibit real-time performance and high accuracy for certain classes of scene. The aim of this paper is to assess the performance of some of these background modelling techniques, namely the Gaussian mixture model and the hybrid detection algorithm, using video sequences of outdoor scenes where the weather introduces unpredictable variations in both lighting and background movement. The results are analysed and reported, with the aim of identifying suitable directions for enhancing the robustness of motion detection techniques for outdoor video surveillance systems.}, }</bibtex>