MEDICAL IMAGE WATERMARKING BY GENETIC ALGORITHM IN LOW FREQUENCY REGION OF IMAGE

Images have important role in different section of life like, social, medical diagnosis, environmental science, etc. Images are easy to modify by different tools and methods, hence reality check or authentication required. Out of many techniques of authentication, invisible watermarking was old and robust. This paper has proposed a robust watermarking model for embedding watermark in medical diagnosis image for patient data validity. Genetic algorithm TLBO was used for finding the DWT coefficient which suits the embedding criteria. Proposed model has embedded the LSB portion of selected pixel coefficients. In this work genetic algorithm increases the randomness of embedding, so at extraction same set of pixel coefficient gives correct watermark information. Other advantage was use low disturbance in carrier image. Experiment was done on real image set of DICOM images of brain tumor for embedding. Results shows that proposed model has increases the robustness of work against the spatial attacks.