Abstract: The requirements needed for an effective and proficient watermarking system is application dependent. However, robustness and image quality (imperceptibility) are fundamental requirements for applications that deal with image watermarking. The major factor that affects the robustness and imperceptibility is the watermark embedding strength. In this study, a CVNN based adaptive technique of estimating watermark embedding strength for a digital image is presented. Experimental results indicated that CVNN based method can estimate the watermarking strength, gives a better correlation and an improved imperceptibility of the watermarked image. It also demonstrates that the detection is enhanced. The use of this new method in watermarking achieved content authentication and helps overcome the problem of visual artifacts and distortions created during watermark embedding.