Watermarking-based Color Image Authentication with Detection and Recovery Capability

In this paper a watermarking-based color image authentication with detection and recovery capability is proposed, in which a Halftone image is used as an approximated version of the luminance channel (Y) and a coded version of the coefficients of the Two Dimensional Discrete Cosine Transform of the chrominance channels (Cb, Cr) of the YCbCr color space are used as watermark signal. The luminance and chrominance information is embedded into the sub-bands of lowest frequency of the Integer Wavelet Transform of the original luminance channel using Quantization Index Modulation Dither Modulation. The luminance information in embedded into the LL sub-band meanwhile the chrominance information is embedded into the LH and HL sub-bands. Besides, a Multilayer Perceptron Neural Network is used as inverse halftoning to increase the quality of the recovered image. The experimental results show the effectiveness of the proposed method compared with some published before.