透過您的圖書館登入
IP:3.133.140.79
  • 學位論文

使用多魚眼攝影機建立環場監控影像

Building Around Video Monitoring System Using Multiple Fish-Eye Cameras and Top-View Transformation Stitching

指導教授 : 林道通
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


在交通安全的議題中,車輛的視野死角已經逐漸被注意,針對駕駛輔助系統已經有許多研究。為解決盲點問題,汽車業者在車輛周圍設置四個攝像機並且建立周圍的影像輔助駕駛。然而,許多這樣的系統仍然有缺陷和盲點存益。本研究涉及通過使用多個魚眼攝影機來建立車輛周圍的360°俯視圖畫面。該系統由三部分組成:魚眼畫面畸變校正,俯視圖像轉換,以及影像合併。我們使用Ration-Function來實現校正畫面,然後再將畫面使用矩陣轉換成俯視圖,並且將矩陣整合以減少誤差。

關鍵字

環場影像 魚眼畫面

並列摘要


Constant innovation in traffic and vehicle safety has prompted attention to the issue of blind spots around the vehicles during driving. Extensive research concerning driver assistance systems has been conducted and the blind areas around a vehicle have been determined to constitute a main threat to traffic safety. To address the problem of blind spots, some manufacturers install four cameras around the vehicle and build a top-view image of the surrounding area that is shown on an interior display. However, many such systems exhibit defects and blind areas remain. This study involved constructing a 360◦ top-view image of a vehicle by using multiple sh-eye cameras. The system consists of three components: distortion calibration of the fish-eye lenses, top-view image transformation, and merging of four images. We used Ration Function to implement the calibration and used matrix to transform the image to top-view, then utilized the property of matrix to reduce the error to complete the 360◦ image.

參考文獻


[1] David Claus and Andrew W Fitzgibbon. A rational function lens distortion model
for general cameras. In IEEE Computer Society Conference on Computer Vision and
[2] Massimo Bertozzi, Alberto Broggi, Paolo Medici, Pier Paolo Porta, and Agneta Sjogren.
Stereo vision-based start-inhibit for heavy goods vehicles. In IEEE intelligent Vehicles
[3] Richard Szeliski and Heung-Yeung Shum. Creating full view panoramic image mosaics

延伸閱讀