Selection of Image Blocks using Genetic Algorithm and Effective Embedding with DCT for Steganography

Abstract

In today's digital world that we live in, security of information is crucial in various communication applications that are widely developed. Steganography is one of the highly secure information hiding techniques. It provides invisible communication and hides the existence of information. This paper focuses on 'before embedding technique' of hiding in image steganography by trying to find suitable places in cover image to embed the secret image. Genetic algorithm (GA) is applied to identify appropriate places in cover image where embedding of secret image will cause minimum distortion. After obtaining these places, embedding is performed using transform domain technique Discrete Cosine Transform (DCT). The secret image is first normalized and then embedded in the lower energy DCT blocks of the selected cover image regions. The experimental results show that the stego images obtained from the proposed method have less visual distortion with satisfactory values in parameters like MSE, PSNR and Correlation used for performance evaluation.