Development of a new approach to improve the performance of steganographic methods in data hiding

Access to data has become very easy thanks to the developing up-to-date technologies. However, the use of these current technologies by malicious people has also revealed security problems. Various techniques have been proposed for reliable storage and transmission of data. In this thesis, two methods have been proposed. These are GA-OPS method and WPT-NS method. In both methods, 20 images of 512x512x3 and 256x256x3 sizes have been used randomly selected from the DICOM library. The similarities of pixels has been used to determine the appropriate pixel positions in data hiding with the GA-OPS method. A genetic algorithm has been used to determine the appropriate data hiding type. After hiding the data, a key has been created by using lifting transform on the location of data hiding and the key has been sent to the other side. According to the results of the analysis, when 1000, 5000 and 10000 characters of data have been hidden in a 512x512x3 image, the PSNR values were 66,5374, 59,4420 and 56,3936, respectively. When 1000, 5000 and 10000 characters of data have been hidden in a 256x256x3 image, the PSNR values were 60,4308, 53,3529 and 47,4113, respectively. In data hiding with WPT-NS method, the image has been divided into sub-bands through 2-level wavelet packet transformation and neutrosophic set approach has been applied. When 1000, 5000 and 10000 characters of data have been hidden, the PSNR values were found 58,1388, 54,5505 and 51,9999 for k=1 and 57,2995, 53,0367 and 50,5468 for k=2, respectively.