Two data hiding / watermarking techniques for grayscale and color images are presented. One of them is DCT based, another uses DFT to embed data. Both methods were implemented in software utilizing C/C++. The complete listings of these programs are included. A comprehensive reliability analysis was performed on both schemes, subjecting watermarked images to JPEG, SPIRT and MPEG-2 compressions. In addition, the pictures were examined by exposing them to common signal processing operations such as image resizing, rotation, histogram equalization and stretching, random, uniform and Gaussian noise addition, brightness and contrast variations, gamma correction, image sharpening and softening, edge enhancement, manipulation of a channel bit number and others. Methods were compared to each other. It has been shown that the DCT method is more robust and, hence, suitable for watermarking purposes. The DFT scheme exhibits less robustness, but due to its higher capacity is perfect for data hiding purposes.