A New Semi-fragile Watermarking for Audio Authentication

This paper presents a new semi-fragile watermarking algorithm for audio authentication, which takes full advantage of scrambing algorithm. Watermark scrambling algorithm can dispel the pixel space relationship of the binary watermark image. After scrambing, the white pixels and black ones of original watermark image are comparatively orderliness, which is robustness against signal processing and fragility to malicious attack. Therefore the proposed scheme can tolerate general signal processing, e.g. MPEG Audio Layer 3 (MP3) compression, and can detect any spiteful tamper on watermarked audio; locate spatial regions from the ruleless of pixels without the help from the original watermark, and even can evaluate the temper degree. Experimental results show that the algorithm can identify intentional content modification and incidental tampering, and also indicate the location where a modification takes place.