Application of adaptive thresholding wavelet filter restoration in steganalysis
自适应阈值小波滤波复原在隐写分析中的应用

When the steganography in spatial domain was modeled as an image degradation process caused by additive noise, the image restoration could be introduced for secret messages detecting. A new method of LSB matching steganalysis was proposed through using adaptive thresholding wavelet filter restoration in steganalysis and combine with the difference characters between the original image and the filtered one. Experimental results show that, the method is better than Ker's calibrating method, especially in detecting uncompressed high resolution raw image due to its simple realization without estimating the embedding rate beforehand.