An adaptive image steganography approach depending on integer wavelet transform and genetic algorithm

Steganography is the art and a science of obscuring the presence of communication by hiding content in electronic media and so obscuring the presence of communication from the adversary’s perspective. The goal of spatial adaptive approaches is to incorporate additional information in the image’s edge regions. The portions of the picture with the most changes are prioritized for embedding in these approaches. In contrast, wavelet-based approaches insert in high-frequency sub-bands to resemble the human visual system. The concept described in this study is a hybrid of these two concepts. On the contrary, the coefficients of the wavelet transform’s high-frequency sub-bands are more suited for embedding, since they have bigger coefficients surrounding them and reflect the image’s edge regions. Based on a local neighborhood analysis, an edge intensity criterion is employed to identify the suitable embedding coefficients in this technique. The receiver may also recognize these coefficients and fully extract the encoded data. In the suggested technique, the picture is first blocked, and then each block is given a wavelet transform. Several coefficients are discovered for each high-frequency sub-band, depending on the duration of the data, and then, using a genetic algorithm and the coefficients are chosen from the detected coefficients using a genetic approach to ensure that the resulting stego picture has the maximum PSNR value. The results of the implementation reveal that in the suggested strategy, employing the Integer Wavelet Transform is far more effective than using the Discrete Wavelet Transform. The suggested approach is safe against steganalysis assaults such as PDH analysis, RS, and universal steganalysers, and the quality of the stego picture is better than previous methods like SPAM (Steganography by Printed Arrays by Microbes) and SRM (Spatial Rich Model).