COVERLESS IMAGE STEGANOGRAPHY FOR PATIENT DATA PROTECTION USING GABOR-DWT FEATURE HASHING

Coverless information hiding has gained increasing interest due to its steganalysis resilience, but existing approaches have the limitation of low capacity and reliance on large datasets. In this approach, a successful coverless image steganography (CIS) technique is proposed with the aim of protecting patient information using one medical image without changing its content. The method has the following steps: Firstly, the confidential message is encrypted using XOR. and Arnold transform and divided into numerous pieces. Secondly, cover medical image is filtered using Gabor filters and Discrete Wavelet Transform (DWT) to yield robust features. Three, a hash table is formed using the resultant features. Four, each piece of the message is mapped to an index in the hash table. Fifth, the indexes are kept in an auxiliary file. Finally, the original image and the auxiliary file are transmitted to the receiver. Experimental outcomes show that the proposed CIS technique achieves zero Bit Error Rate (BER = 0) and 100% Recovery Correctness (RC) and is able to achieve high capacity along with immunity to common image processing attacks. Compared to previous techniques such as the technique employed in [22], which requires more than 550 medical images for scanning all alphanumeric characters, the new method functions as equally or even superior with a single medical image, a significant advancement in terms of efficiency, scalability, and usability.