Finding the area fraction of ink in a halftone image is an important part of judging the quality of the image produced and the quality of the system that produced it. In order to improve histogram analysis from mere visual judgment, algorithms were constructed to come up with a function that would computationally model a graphical representation of a histogram that could match the histogram of a real halftone pattern. The variables within the model, in turn, gives the desired area fraction of ink value. Two different algorithms were developed. In one, the histogram is treated as two Gaussian functions and the model is the summation of fractions of the two functions. The other uses the equivalent straight edge of the halftone pattern. The model here is the derivative of the function of the edge convolved with a Gaussian noise metric. By finding the minimum root mean square deviation and least variances of the difference between the model and actual data across the gray levels of reflectances, a good estimate of the area fraction of ink F and other quality metrics of the histogram of a halftone pattern could be made.