Adaptive Threshold Based-Enhanced Bit Plane Complexity Segmentation for High-Embedding Capacity Image Steganography

This study developed an Adaptive Threshold-Based Enhanced Bit Plane Complexity Segmentation (AT-EBPCS) technique for high-capacity image steganography. Conventional BPCS methods use fixed complexity thresholds, limiting adaptability to images with varying texture distributions. The proposed technique employs an adaptive thresholding mechanism to dynamically identify suitable embedding regions based on local image characteristics. Implemented in MATLAB R2023a using thirty (30) Chest X-ray images, the technique achieved an average PSNR of 50.75 dB and payload ratio of 23.48%, outperforming Chandra Sekhar et al. (2018) at 42.00 dB and 21.00%, and Nwankwo et al. (2023) at 38.00 dB and 19.20%. Shannon entropy increased from 7.4168 to 7.5288, indicating improved randomness and statistical security. The adaptive thresholding mechanism significantly enhanced embedding capacity, visual imperceptibility, and statistical resistance, making AT-EBPCS suitable for secure high-capacity image steganography.