Biometric image-based Image steganography and steganalysis using EQC-Net

In this paper, an efficient quantum convolutional network (EQC-Net) based framework for image steganography and image steganalysis is proposed. In the process of steganography, initially, a bitmap image is obtained from the input biometric cover image. Later, the secret message and key are XORed and applied along with the bit map image for generating the embedded image to different embedding techniques, such as discrete cosine transform (DCT), discrete wavelet transform (DWT), and least-significant bit (LSB)-based embedding. After the steganography process, the determination of the hidden message is done using steganalysis. To detect the secret information the embedded image is converted into a bitmap image by utilising the EQC-Net hybrid model. The evaluation of the EQC-Net shows that the bit error rate (BER) and peak-signal-to-noise ratio (PSNR) achieved by the EQC-Net is 4.887 and 46.61 6dB for LSB-based embedding.