Digital images have many uses in the field of health, research, military, art, etc. Digital Images need annotation for retrieval and protection from piracy, attacks, and modification. To perform this retrieval and protection, some of the information was hidden in the image matrix. This paper proposes a data embedding and extraction algorithm. Low-frequency regions of the image are identified to embed secret information in LSV (Least Significant Value) and MSV (Most Significant Value) of a selected coefficient once the data hiding in the embedding section of the proposed model image gets retransformed in the original image structure as per the number of bits used in LSV and MSV for data hiding image gets secured from different spatial attacks. The experiment is performed on real images. A comparison of the proposed model is done on PSNR, MSE, evaluation parameters. It is shown that the proposed model is better as compared to the existing model.