Increase in research and commercial interest is observed in the area of digital watermarking in current era. Main reason behind this development is the extensive use of internet with high speed and bandwidth availability. As a result, sharing of multimedia content is increased significantly, which rises need for copyright protection and authentication. Digital watermarking provides vast commercial applications for securing content of different multimedia objects. However, watermarking can cause permanent loss in the content of the object. Different sensitive multimedia objects like medical, military, scientific etc. do not tolerate such permanent loss. Therefore, the distortion needs to be minimized or completely eliminated so that the true purpose of the multimedia content can be retained. Numerous watermarking techniques are available for different sensitive applications. Our objective is to propose reversible and secure techniques for watermarking different multimedia objects. Robust watermarking approaches are targeted in this research. Main focus is to achieve maximum watermarking capacity and to minimize embedding distortion, while preserving functional capability of the underlying multimedia object. Two different types of multimedia objects are targeted in our dissertation; these include relational databases and DNA medium. Genetic algorithm is selected in order to apply intelligent techniques for improving different properties of digital watermarking. Two relational database watermarking approaches are designed and developed; both approaches are reversible, robust, and follow the distortion tolerance of the attributes. First approach uses genetic algorithm to improve capacity, reduce distortion, and false positive rate of the difference expansion based watermarking technique. Whereas, second approach improves watermarking capacity, reduces distortion, and false positive rate of the reversible contrast mapping, which is first time designed for relational databases. A robust data hiding technique for DNA medium is proposed, which increases watermark capacity by improving current synonymous substitution technique and resists mutation losses. Moreover, synonymous substitution does not causes any disturbance in the amino acid sequence, thus the DNA functionality is retained. In order to tackle different DNA mutations binary strings and Reed Solomon Codes are applied. The watermark is encoded before embedding using Reed Solomon Codes and the structural information is retained using binary strings.