Contrast Modification Forensics Algorithm Based on Merged Weight Histogram of Run Length

Contrast Enhancement Detection Based on Histogram

Stamm et al. (2010) have proposed an algorithm of contrast enhancement detection based on histogram peak/gap artifacts left by attacked image; the main steps are described in the following:

• 1.

  Calculate the image’s pixel value histogram and the modified histogram such that:

  
  (1)

where, is a pinch off function, whose role is to eliminate the low end or high end saturated effect in image.

• 2.

  Calculate the high-frequency measurement F according to the following formula:

  
  ,
  
  (2)

where, N is the total number of pixels, is the discrete Fourier frequency transform of , and is the cutoff function deemphasizing the low frequency components of G(k):(3) where, T corresponds to a desired cutoff frequency.

• 3.

  At last, for a given threshold , F is compared with the to determined whether the image has been modified.