The method used to conceal data in various formats, including text, photos, music, and video, is called steganography. A number of techniques are being developed to conceal sensitive information in photos, including the widely used technique of image steganography. Although they are highly imperceptible and have a large hiding capacity, conventional steganography techniques like LSB and DCT are not secure. This thesis proposes a novel method that applies singular value decomposition (SVD) to obtain singular values and Discrete Wavelet Transform (DWT) to obtain wavelet coefficient on the cover picture. A non-dominated categorization genetic technique is worn to embed the secret image into these solitary values after it has been jumbled using chaos. The primary objective of the method that is being given is to enhance the stego-image’s representation excellence and embedding capacity. The algorithm is made more secure with the use of scrambling. In terms of PSNR, the experimental findings exhibit that the suggested technique performs improved than the similar methods. Both of the quantity of data covered and the standard of the cover picture are taken into consideration when estimating the model’s efficacy.