Tamper detection and recovery of medical image watermarking using modified LSB and Huffman compression

Abstract

Medical image is an object which can help doctors to diagnose the patient. However, due to the rapid development of technology nowadays, mostly there are several parties that did some manipulation towards the medical image by using this technology. It can be harm for the patient. One of the concrete solutions to protect the authenticity of an image from the manipulation is by applying a watermarking. By inserting the watermark, the process to detecting the location of a manipulated medical image can be done by this system. Hence, the manipulated medical image can also be recovered until it similar with the original image by inserting a feature extraction such as the average intensity of the image blocks. This paper will discuss a scheme of reversible watermark using a modified LSB and Huffman compression to detecting and recovering the manipulated medical image. The testing results shows that this system is capable to detect the attack with an accuracy of up to 100% and can do the recovery with an accuracy of recovery rate up to 98% for some attacks.