Research And Implementation Of Advertising Detection System Based On Hadoop Platform

The fusion of telecom network, Internet and broadcasting network is an important direction of Chinese information industry development, and the fusion of the Internet and broadcasting network is one part of it. With the development of the communication industry, the Internet has the ability of transferring video, audio data with high quality, and the transmission mode is bidirectional. So if the fusion is complete, the TV user can have the unprecedented interaction way using the bidirectional feature.As the " Twelfth Five-Year Plan" and the State Council for speeding up the cultivation and development of strategic emerging industries," cloud computing " has become a new generation of information technology industry is the important part of. The integration and combination of "cloud computing " Internet data transmission is high speed bidirectional characteristic and cloud computing speed stability computational advantages complement each other, can give full play to each other’s advantages.For the interactive television playback function, in the television automatically remove the advertising content will undoubtedly improve user interaction experience. This paper presents a method based on video commercial detection algorithm. As a result of the process to deal with massive data, so this design uses current cloud computing platform Hadoop with suspected advertisement detection algorithm based on video image processing, on the image by using machine learning methods, and improves the system performance. Also for the cloud computing platform for image processing to do try.