Image steganography is the process of hiding secret information behind the cover image. This paper employs the graph wavelet method for image steganography. In the proposed work, the high-quality stego image is provided by the inter-pixel relationship management function of graph wavelet. A scrambling technique based on the pixel intensity of a 2D image is used to raise the secret image's degree of invisibility. Image extraction is the second goal, and it is accomplished using graph signal processing (GSP). Because it saves the neighborhood pixel information, GSP provides an extracted image of higher quality. In the section on experimental results, the value of the blending coefficient is changed to improve the look of both the stego image and the secret image that was extracted from it. For the suggested technique between cover and stego images, the peak signal-to-noise ratio (PSNR) ranges from 44.99 to 37.72 dB for different alpha values. The parameters of the naturalness image quality evaluator (NIQE) is used to analyze the secret image that has been extracted. For the extracted secret image, the NIQE observer score is 3.49. The proposed work tested against a dataset from a previously published paper. The resulting stego’s are better and acceptable.