Dynamic Steganography Least Significant Bit with Stretch on Pixels Neighborhood

Abstract

Background: The confidentiality of a message may at times be compromised. Steganography can hide such a message in certain media. Steganographic media such as digital images have many pixels that can accommodate secret messages. However, the length of secret messages may not match with the number of image pixels so the messages cannot be inserted into the digital images. Objective: This research aims to see the dynamics between an image size and a secret message’s length in order to prevent out of range messages entered in an image. Methods: This research will combine the Least Significant Bit (LSB) method and the Stretch technique in hiding secret messages. The LSB method uses the 8 th bit to hide secret messages. The Stretch technique dynamically enlarges the image size according to the length of the secret messages. Images will be enlarged horizontally on the rightmost image pixel block until n blocks of image pixels. Results: This study compares an original image size and a stego image size and examines a secret message’s length that can be accommodated by the stego image, as well as the Mean Square Error and Structure Similarity Index. The test is done by comparing the size change of the original image with the stego image from the Stretch results, where each original image tested always changes dynamically according to the increasing number of secret message characters. From the MSE and SSIM test results, the success was only with the first image, while the second image to the fourth image remained erroneous because they also did not have the same resolution. Conclusion: The combination of LSB steganography and the Stretch technique can enlarge an image automatically according to the number of secret messages to be inserted. For further research development, image stretch must not only be done horizontally but also vertically.