Projeto de esteganografia computacional: construção e validação de um sistema de ocultação de mensagem em python

Abstract

This work presents the design, implementation and validation of a computational steganography system developed in Python, focused on the secure concealment of textual messages within digital images. The system integrates two widely studied steganographic techniques - LSB (Least Significant Bit) and BPCS (Bit-Plane Complexity Segmentation). To evaluate the performance of the implemented techniques, the metrics PSNR (Peak Signal-to-Noise Ratio), SSIM (Structural Similarity Index), and bitrate were applied to images at three distinct resolutions - low (74×131 pixels), medium (368×656 pixels), and high (1472×2626 pixels) - enabling a comprehensive and systematic comparative analysis. The results demonstrated that both techniques are capable of concealing information with satisfactory visual imperceptibility. The LSB method showed greater structural preservation of the image, while BPCS offered greater storage capacity per block. The obtained metrics confirmed that the higher the resolution of the cover image, the lower the visual impact of the modifications, highlighting the importance of proper sizing of the cover medium for steganographic applications.