由於駕駛車輛安全性問題，駕駛輔助系統相關技術在距今約二十多年前開始被重視與開發，透過攝影機架設在車輛內以視覺式的方式分析車輛前方道路的狀況來輔助駕駛人。其中路面偵測技術在先進駕駛輔助系統中扮演重要的角色，因為該技術不僅能提供正確的路面區域、道路形狀及標線位置，還能將諸多前方道路資訊提供給其他的駕駛輔助系統使用。
路面偵測技術已經逐漸的成熟，但是較少研究者針對雨天偵測路面系統進行開發與研究。透過整理近年的路面偵測文獻發現，幾乎所有研究都會先透過邊緣偵測相關技術讓系統得知影像中何處是正確的路面區域，以便進行路面顏色特徵擷取並執行路面偵測。而也有部分的文獻是先直接手動給予路面區塊，進行路面顏色特徵的擷取並初步的偵測出路面，再利用邊緣偵測技術進行改善。可以發現邊緣偵測技術對於找出路面位置有莫大的幫助，但是一旦道路標線或路面邊界模糊不清時，便無法達到效果。
因此本研究提出的雨天路面偵測系統主要是透過log chromaticity space、sensor sharpening matrix transform以及watershed segmentation來進行雨天路面偵測，不需要透過邊緣偵測技術且會自動採樣合適的路面區塊作為系統訓練的依據。首先，系統一開始時必須先確定自動擷取的region of interest (ROI)是路面區域，接著才進行road feature extraction，透過執行log chromaticity space、sensor sharpening matrix transform及projection and thresholding各步驟後便可以將整張影像與ROI具有相似顏色的像素視為同一類，最後再結合watershed segmentation影像分割技術來改善誤判。
最後實驗的部分，本研究針對小雨、中雨、中大雨三種雨勢以及數種道路形狀進行實驗，實驗結果呈現出，對於較遠的車輛、較大的雨勢以及過度曝光的路面經常會產生誤判，但若是實驗場景中有著充足的光線以及道路上的車輛都有色彩鮮明的顏色，路面偵測的結果都能產生很高的正確率。

The last two decades, for the driving safety, in-car video camera technology for estimating road shape ahead of a vehicle for the purpose of driver assistance has been developed and implemented. A road detection technology can provide the road shape, area and lane markings, which can share with other advance driver assistance systems (ADAS).
Although road detection is a mature technology, there are a few studies on road detection for rainy days. Recent studies use mostly road boundary to detect road area. If road boundary is not clear, it would fail to detect the correct road area.
To effectively handle road detection on rainy days, the proposed method comprises three modules: 1) log chromaticity space, 2) sensor sharpening matrix transform, and 3) watershed segmentation. To speed up the road detection processing, the region-of-interest (ROI) is used to choose the candidate searching range for extract road features. The system then process the log chromaticity space method and sensor sharpening matrix transform method. Finally, the watershed segmentation method as an error correction mechanism is applied.
Experimental results on real road scenes such as light rain, moderate rain, and heavy rain show that if road scenes is bright enough and vehicles have vivid color, the proposed method has substantiated the effectiveness of accuracy results.

[Alv11] J. M. Alvarez and A. M. Lopez, “Road Detection Based on Illuminant Invariance,” IEEE Transactions on Intelligent Transportation Systems, vol. 12, no. 1, pp. 184-193, 2011.
[Chi05] K. Y. Chiu and S. F. Lin, “Lane detection using color-based segmentation,” in Proceeding of IEEE International Conference on Intelligent Vehicles Symposium, Nevada, USA, pp. 706-711, 2005.
[Foe04] M. Foedisch and A. Takeuchi, “Adaptive Road Detection through Continuous Environment Learning,” in Proceeding of IEEE International Conference on Information Theory, Chicago, USA, pp. 16-21, 2004.
[He04] Y. He, H. Wang, and B. Zhang, “Color-Based Road Detection in Urban Traffic Scenes,” IEEE Transactions on Intelligent Transportation Systems, vol. 5, no. 4, pp. 309-318, 2004.
[Hon02] T. H. Hong, C. Rasmussen, T. Chang, and M.Shneier, “Road Detection and Tracking for Autonomous Mobile Robots,” in Proceeding of IEEE International Conference on Aerospace/Defense Sensing, Simulation, and Controls, Orlando, USA, pp. 311-319, 2002.
[Hu04] M. Hu, W. Yang, M. Ren, and J. Yang, “A Vision Based Road Detection Algorithm,” in Proceedings of IEEE International Conference on Intelligent Vehicles Symposium, Singapore, pp. 846-850, 2004.
[Kim11] B. Kim, J. Son, and K. Sohn, “Illumination Invariant Road Detection Based on Learning Method,” in Proceedings of IEEE International Conference on Intelligent Transportation Systems, Washington, USA, pp. 1009-1014, October, 2011.
[Rot04] C. Rotaru, T.Graf, and J. Zhang, “Extracting Road Features from Color Images using a Cognitive Approach,” in Proceedings of IEEE International Conference on Intelligent Vehicles Symposium, Parma, Italy, pp. 184-191, 2004.

[Wan08] Y. Wang, D. Chen, and C. Shi, “Vision-Based Road Detection by Adaptive Region Segmentation and Edge Constraint,” in Proceedings of IEEE International Conference on Intelligent Information Technology Application, Shanghai, China, pp.342-346, 2008.
[Yen13] S. Yenikaya, G. Yenikaya, and E.Duven, “Keeping the vehicle on the road: A survey on on-road lane detection systems,” ACM Transactions on Computing Surveys, vol. 46, no. 1, pp. 1-43, 2013.
[Bmw11] BMW官方網站，取自http://www.bmw.com.tw/home/index.htm，2011年。
[Car10] Goo新車車訊網，取自http://www.carnews.com/news/article/19499，2010年。
[交10] 交通部中央氣象局。氣候統計。取自http://www.cwb.gov.tw/V7/climate/monthlyMean/Taiwan_precp.htm，2010年。
[聯15] 聯合新聞網。新聞。取自http://udn.com/news/story/7087/905748-Google%E7%84%A1%E4%BA%BA%E8%BB%8A-%E5%A4%8F%E5%AD%A3%E9%96%8B%E4%B8%8A%E5%85%AC%E7%94%A8%E9%81%93%E8%B7%AF%E5%AF%A6%E6%B8%AC，2015年。
[維14] 維基百科。圖像分割。取自https://zh.wikipedia.org/wiki/%E5%9B%BE%E5%83%8F%E5%88%86%E5%89%B2。