A system for transforming coordinate systems in an automated video monitor alignment system includes a camera and computer to capture an image of the display bezel surrounding the cathode ray tube and of the image displayed on the CRT. Three dimensional models of the CRT tube and display bezel are used to compute and compensate for camera perspective. Camera pixel data can then be transformed to `flat plane` coordinates. When the relative orientation of the camera and CRT are corrected, the effects of parallax can be eliminated, allowing more accurate inspection, measurement, and lower cost fixturing on the production line. Thus, an accurate coordinate transformation from a camera-referenced coordinate system to a monitor-referenced coordinate system is made. Correction factors for correcting errors due to refraction can also be generated empirically in accordance with the present invention. Slope coordinate data is generated for a sample monitor for predetermined slope coordinate locations on the tube face. Correction signals can be generated from the slope coordinate data and measured tilt angles of the monitor relative to the vision system. Two dimensional interpolation techniques can be employed to provide slope coordinate values and correction signals for video signals that do not physically match the location of the slope coordinate locations of the sample monitor.