Monitor Gamma Correction

In this part of the project, we try to estimate the value of the display intensity which is also known as the display gamma value in order to calibrate our display. Calibrating a visual display means measuring the relationship between the video frame-buffer values and the intensity of the light emited by the display so that we are able to control the appearance of the display. In a color CRT display, there are three separate video frame-buffers each with a matrix that specifies the control signal for one of the electron guns that in turn stimulate one of the three phosphors. Each matrix entry specifies the desired voltage level of the control signal point on the faceplate. Experiments have shown that the relative intensity of the light emitted from the CRT display I and the frame-buffer value V re related by the following equation: I = a.(I)^gamma + b where a and b are two fitting parameters. We assume a simpler relationship between I and V in our analysis: I = V^gamma.

Our procedure in determining the value of gamma for our monitor (all the testing was done on an HP A4331A monitor with the brightness level set equal to 30 and the contrast set equal to 100) was as follows:

1. In matlab, we created two matricies Alt_Mat and Const (both 128x128 pixels) where Alt_Mat consists of alternating values of 0 and 254. Const was set to the initial costant value 128.

2. We displayed the two matricies as images using the matlab command imshow(matrix_name) .

3. Using the Colorometer, we compared the two CIELAB Luminance values (Y-values) for the two images.

4. The Alt_Mat was kept constant while we tried to vary the value of Const to match the luminance level of Alt_Mat. The best match was found to be Const = 197 with the help of the colorometer. The two images are shown below for illustration:

5. The value of gamma was calculated from the equation:

I = V^gamma

by equating the frame-buffer values for the two images.

[(254)^gamma + (0)^gamma]/2 = (197)^gamma

=> gamma ../ 2.7

I wrote the following matlab script to estimate the value of the monitor gamma by carrying out the above steps.

Please refer to it for further details.

gammaCorrection.m


The above explanation of gamma correction was adapted from "Foundations of Vision" by Brian Wandell pp 414-416.

<=Previous Page

Back to HOME