Chronological bald eagle optimization based deep learning for image watermarking

This work presents a Chronological Bald Eagle Optimization (CBEO)-Deep Learning (DL) approach for performing image watermarking. Here, the watermark is implanted in the cover image, by selecting the optimal region in the cover image with the help of the LeNet. Further, the Haar Wavelet Transform (HWT) is utilized in the embedding procedure to improve the robustness of the approach. The trainable parameters of the LeNet used for selecting the optimal region in the cover image are optimized utilizing the CBEO algorithm. Furthermore, the effectiveness of the HWT + CBEO_LeNet is inspected by considering parameters, such as Normalized Correlation (NC)and Peak Signal-to-Noise Ratio (PSNR), and investigations reveal that the proposed HWT + CBEO_LeNet offered high robustness against various noises and attacks and computed a maximum PSNR, NC, and SSIM of 24.989 dB 0.761, 0.969 and obtained least BER value of 0.047, respectively.