A new Crypto-Stegano system is proposed using Random Matrix Transformation on relations of association schemes based on finite abelian groups associated with Arnold tranform and concealing the encrypted data within a randomly selected cover image. The proposed technique has been evaluated through comprehensive experimental results. Additionally, statistical tools are used to assess its performance. Furthermore, error analysis and key analysis have been carried out to ensure the technique’s accuracy and reliability. Comparison results between existing schemes and the proposed techniques to validate the robustness have also been provided.