An adaptive steganographic algorithm for 3D polygonal models using vertex decimation

Most 3D steganographic algorithms emphasize high data capacity, low distortion, and correct data extraction. However, their disadvantage is in the existence of the same embedding capacity for each data-embedded vertex in the 3D models. Embedding the same capacity in the vertex located on the surface with different properties may cause obvious distortion, making it difficult to achieve the initial goal of information-hiding techniques. This study proposes a new and adaptive 3D steganographic algorithm that considers the surface complexity. To increase the accuracy of the complexity estimation for each embedding vertex, the proposed algorithm adopts a vertex decimation process to determine its referencing neighbors. Thereafter, different amounts of the secret messages are embedded according to the surface properties of each vertex. This approach preserves important shape features and produces a more imperceptible result. Experimental results show that the proposed adaptive algorithm can achieve more accurate estimation results with a higher data capacity and acceptable distortion. The proposed technique is feasible in 3D steganography.