Position estimation using audio digital watermarking for prevention of movie theft in theaters

Recently, copyright violation due to movie piracy becomes a serious problem. According to the report from Motion Picture Association of America, the annual loss due to the movie piracy exceeds six billion dollars. Many of pirated movies which become available soon after premier of the movies come from movie theft in theater, i.e., a movie on screen is recorded with a camcorder. The movie theft in theater is banned by the law in many countries, and the countermeasure against this problem is strongly required.

In this research, we propose a system for estimating the position of the camcorder which is used for movie theft using audio digital watermarking technique. In the proposed system, a model of watermark which undergoes the propagation in the air is constructed and the position is estimated based on the model. The proposed system achieves high estimation accuracy using the assumption that a movie is recorded at the same position for a long duration. Furthermore, the proposed system is able to estimate the position without modifying the existing acoustic facility of the theater if the theater has more than three loudspeakers. We expect that the proposed system can be a powerful deterrence against the movie theft in theater.

Real-time position estimation using audio digital watermarking

In this research, aiming at application to location-based services such as navigation, we propose a system for estimating a user position in real-time. In the proposed system, assuming that multiple loudspeakers are installed in the environment, music into which a watermark is embedded and which is emitted from each loudspeaker is received by a mobile device. Then, the position of the mobile device is estimated using a model of the watermark which undergoes propagation in the air. Since we cannot assume that the music is received at the same position for a long duration as in the position estimation for movie theft deterrence, we propose a new model of the watermark which is more sophisticated than that for movie theft deterrence and adopt the particle filter in order to improve the estimation accuracy.

Recoverable privacy protection using information hiding technique

When a surveillance video is distributed on the Internet, subject's region should be hidden by using image processing techniques for protecting subject's privacy. However, under some situations such as criminal investigations, and monitoring children, it is strongly required to recover the original image from the privacy protected image. We refer to such privacy protection as recoverable privacy protection.

For recoverable privacy protection, we have proposed a method using information hiding technique together with discrete wavelet transform (DWT). DWT generates scaling coefficients and wavelet coefficients. Scaling coefficients, which can be regarded as a low resolution image of an original image, are used for producing privacy protected image. Wavelet coefficients, which can be regarded as privacy information, are embedded into the privacy protected image via information hiding technique. Therefore, only authorized viewers can extract privacy information and recover the original image from the privacy protected image.