Abstract: To enhance security of steganography, the characteristics of the image should be kept as far as possible. This paper proposes a steganography algorithm with histogrampreserving property. Firstly, the histogram of pixel values is grouped. By calculating the maximum payload under a given distortion, the modification rate for each pixel is found and the location of pixels is got that need to be changed by syndrome trellis codes (STC). Then embedding is performed by changing those pixels using adaptive intra-group and/or inter-group adjustments. The difference in histogram between the cover and the stegoimages is very small. Experimental results show that the proposed method makes less histogram change and gives better visual quality than existing methods under the same
embedding rate.

Key words: visual quality, histogram, steganography, syndrome trellis code (STC)