Spatial HOG based TV logo detection

Most existing TV logo detection methods use temporal correlation of multiple consecutive frames. However, these methods cannot meet the need of real-time system. This paper considers a robust detection method based on single frame. Our TV logo detecting method takes the spatial information of all matching patches as a whole, and presents a new descriptor SHOG (Spatial HOG) to formulate the spatial distribution. It is compact but informative. The proposed geometry voting method achieves fast locating TV logo with geometry verification in the step of detection, which is also competent for the detection of semi-transparent and hollow TV logos. Experiments show that SHOG is robust to occlusion and small deviation of the key points' position and our proposed method outperforms the state-of-art local feature based template matching method for logo detection as performance and speed concerned.

[1]  N. Otsu.  A threshold selection method from gray level histograms , 1979 .

[2]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[3]  Baoxin Li,et al.  Automatic detection of replay segments in broadcast sports programs by detection of logos in scene transitions , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[4]  Shih-Hsuan Yang,et al.  An improved automatic commercial detection system , 2011, 2011 Visual Communications and Image Processing (VCIP).

[5]  Keisuke Takada,et al.  3D Object recognition using a voting algorithm in a real-world environment , 2011, 2011 IEEE Workshop on Applications of Computer Vision (WACV).

[6]  Ebroul Izquierdo,et al.  Logotype detection to support semantic-based video annotation , 2007, Signal Process. Image Commun..

[7]  Sheng Tang,et al.  Logo detection based on spatial-spectral saliency and partial spatial context , 2009, 2009 IEEE International Conference on Multimedia and Expo.

[8]  George Hripcsak,et al.  Technical Brief: Agreement, the F-Measure, and Reliability in Information Retrieval , 2005, J. Am. Medical Informatics Assoc..

[9]  André Kaup,et al.  Automatic TV logo removal using statistical based logo detection and frequency selective inpainting , 2005, 2005 13th European Signal Processing Conference.

[10]  Pinar Duygulu Sahin,et al.  Comparison and combination of two novel commercial detection methods , 2004, 2004 IEEE International Conference on Multimedia and Expo (ICME) (IEEE Cat. No.04TH8763).

[11]  Yongdong Zhang,et al.  Hollow TV logo detection , 2011, 2011 18th IEEE International Conference on Image Processing.

[12]  Christopher G. Harris,et al.  A Combined Corner and Edge Detector , 1988, Alvey Vision Conference.

[13]  Adnan Amin,et al.  Real-time Detection of Semi-transparent Watermarks in Decompressed Video , 2007, 2007 IEEE Workshop on Applications of Computer Vision (WACV '07).

[14]  Alex Santos.  Real-Time Opaque and Semi-Transparent TV Logos Detection , 2006 .