無線感測網路(Wireless Sensor Networks，WSNs)至今已有許多發展與應用，漸漸融入在我們的生活中，其未來性是不可小覷的。在入侵偵測方面已經有許多相關的研究，但大多研究都只注重在是否偵測到入侵者，而非在辨別入侵者方面。無線攝影機感測網路(Wireless Camera Sensor Networks，WCSNs)中雖利用攝影機去追蹤目標物，但其卻不能保證必定能辨識到對方長相，由於目標物行走方向與攝影機照射角度的不同。此外攝影機的部署成本也遠大於多數的感測器，使用隨機部署方式將可能造成成本的昂貴。因此本篇論文主要是想改善無線攝影機感測網路的部署方式，利用較少的攝影機，且能確保目標物至少被辨識一定的程度。並希望透過最後的分析結果，驗證本篇論文的部署方式，可以在大場景中能有比較少的成本且能有一定的辨識度。 Wireless Sensor Networks (WSNs) has been a lot of development and application. It gradually integrated in our lives, the future is not to be underestimated. In terms of intrusion detection has been many related research, but most studies only focus on whether to detect intruders, rather than in terms of identifying intruders. In wireless camera sensor networks (WCSNs), although we can use cameras to track the target, but it can not guarantee to be able to recognize each other looks, due to the traveling direction of the target with irradiation angles of camera are different. In addition the camera sensor deployment cost is much larger than most other sensors, the sensors random deployment may result in expensive costs. Therefore, this paper mainly to improve the deployment of wireless camera sensor networks, the target can identify at least some extent by using fewer cameras. Finally, through the final results of the analysis, validation of this thesis deployment, in large environments can have relatively little cost and can have a certain level of recognition.
