This paper proposes a Secure, robust and intelligent watermarking approach based on Redundant Wavelet Transform, Singular Value Decomposition and Genetic Algorithm. Here the robustness. The proposed watermarking scheme accomplishes RWT and SVD for feature extraction and the GA for optimization. Further this approach proposes a signature embedding mechanism by which the watermarked image will be more secure. The Genetic Algorithm proposed in this approach adopts intelligent property and the signature process adopts more security. Further the combination of RWT and SVD adopts the robustness to the proposed approach. Numerous experiments are conducted over the proposed approach and the performance is evaluated through the performance metrics such as Peak Signal to Noise ratio (PSNR) and Normalized Correlation (NC). The obtained experimental results demonstrates that the proposed approach is superior compared to conventional approach and is able to provide efficient resistance over Gaussian noise, salt & pepper noise, median filtering, cropping, rotation, contrast enhancement, scaling and Histogram Equalization attacks.