Enhancing the performance of watermarking based on Cat Swarm Optimization method

In this paper, an attempt has been made to retrieve watermark same as original watermark using Cat Swarm Optimization (CSO) technique. Embedding watermarks in frequency domain can usually be achieved by modifying the Least Significant Bits (LSBs) of the transformation coefficients. The hidden watermark in an image is retrieved differently from the original watermark from the frequently used rounding approach. The simple rounding approach cause rounding errors in the process of transformation of image from frequency domain to spatial domain. Swarm intelligence techniques are proposed to correct these rounding errors. The result demonstrates that the computation time and number of iteration taken by the CSO process was also found to be less than PSO and PSO time-varying inertia weight factor method.