以標誌為基礎之即時電視廣告偵測技術

Real-Time TV Commercial Detection Based on Logos

陳宥成(Yu-Cheng Chen)

指導教授 ： 楊士萱

國立臺北科技大學/電資學院/資訊工程系研究所/碩士(2010年)

https://doi.org/10.6841/NTUT.2010.00371

摘要

觀賞電視節目是多數人工作之餘的主要休閒方式，可是在使用電視棒（卡、盒）錄製影片時，廣告的部份通常會一起保留下來，影響到觀賞節目的順暢。許多國家（包括台灣）的電視影片，在廣告播出時節目名稱的標誌(logo)會消失，因此可以做為廣告偵測的依據。本論文提出一套以節目名稱標誌為基礎的廣告偵測方法，改進廣告與節目邊界的判斷，而達到即時廣告偵測的效果。在標誌取得的部份，我們取得影片前數分鐘畫面中不變的顏色，利用保留下來的部份產生標誌的遮罩，再對於取得標誌中佔最高比例的3個顏色，統計其空間分佈情況做為特徵。在標誌比對的部份，我們比對取得標誌與待測標誌的3個主要顏色比例和空間分佈，當誤差大於一個設定的門檻值則視為不相同的標誌。實驗結果顯示，我們提出的方法可以偵測出測試影片所有的廣告片段，而且偵測出來的準確率平均可達90％以上。

關鍵字

廣告偵測 ； 場景轉換 ； 標誌偵測

並列摘要

Watching TV is the mainly entertainment when people have leisure time. When we record the TV programs by video recording devices, the unwanted information may include in TV program, like commercials. The TV programs' logo or TV station's logo will disappear during commercial. Therefore, we can use this feature for commercial detection. In this paper, we propose a real-time commercial detection method based on program's logo. For logo extraction, we used the pixels which have fixed color in the first several frames to get the logo mask. Using 3 dominant colors in logo to get it's spatial distribution. For logo matching, we used logo mask to produce candidate region compared with logo by the spatial distribution of dominant colors. If the differences between two distributions is over the threshold, we treat they're two different logos. After experiment, our method can find out all commercial segments, and the average accuracy can achieved more than 90%.

並列關鍵字

Commercial detection ； shot change ； logo detection

參考文獻

[1] J.H. Yeh, J.C. Chen, J.H. Kuo, and J.L. Wu, “TV Commercial Detection in News Program Videos,” IEEE International Symposium on Circuits and Systems, pp. 4594 - 4597, 2005.

[2] R. Lienhart, C. Kuhmunch, and W. Effelsberg, “On the Detection and Recognition of Television Commercials,” IEEE Conf. on Multimedia Computing and Systems, pp. 509-516, Jun. 1997.

[3] P. Duygulu, M.-Y. Chen, and A. Hauptmann, “Comparison and Combination of Two novel Commercial Detection Methods,” CIVR'04, Jul. 2004.

[5] J. Gauch and A. Shivadas, “Identification of new commercials using repeated video sequence detection,” ICIP, pp. 1252-1255, 2005.

[6] X.-S. Hua, L. Lu, and H.-J. Zhang, “Robust learning-based TV commercial detection,” ICME’05, pp.149-152. 2005.