Coding techniques for data hiding in images

This paper presents the application of source and channel coding techniques for data hiding in still images. The hidden data is source coded by vector quantization, and the indices obtained in the process are embedded in the host signal using discrete wavelet transform and channel coding. The objective of this paper is to evaluate the improvement that channel codes bring to the embedding algorithm. The source coded signature data are channel coded using convolutional and concatenated codes before embedding in the host coefficients. The hidden data are recovered using is decoding which implements the Viterbi decoding technique. The algorithm is general and can be applied to different types of host data. The host signal with the embedded data is compressed before the retrieval of the hidden information. The extracted data are shown for varying compression ratios of the host signals