Browse Prior Art Database

System and method for real-time projector image transforms Disclosure Number: IPCOM000235669D
Publication Date: 2014-Mar-19
Document File: 2 page(s) / 49K

Publishing Venue

The Prior Art Database


Image projectors are frequently used for the display of information on a screen. This projected image is often of trapezoid form, not rectangular, since the project and screen are not always perfectly aligned. To correct for this, the projector and/or projected image surface can be moved and rotated, but often not effectively. This publication describes a method of automatically aligning and transforming the projected image, such that manual intervention is not required.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 2

System and method for real-time projector image transforms

One or more cameras and an image processing unit are integrated into the projector (the embodiment described below uses two cameras). Input from an external source (e.g., laptop) runs through the processing unit before being passed to the usual system for projection. The cameras detect the shape of the projected image on screen, so that the system can calculate the skew, rotation, etc. of the projected image. An inverse matrix or similar calculation is then performed, such that when applied to such a skewed image, it will return to being normal and rectilinear, as originally intended. This transform is now applied to the signal from the laptop for the next several seconds, before being passed to the projection system, such that the projected image appears to be as intended.

    With most projectors, the input signal (assumed to be rectilinear throughout this disclosure) routes directly (through a basic image processing chipset) to the lens, where the image is projected onto a screen. With the technique described in this publication, every nth frame of the input signal is stored in memory in the processing unit in the figure below. The signal is passed on to the lens as usual, to project the image. At this point, the cameras view the projected image and anything in front of the projector, and passes it back to the processing unit.

    Image processing is then performed on this image, to determine the part of the image with the highest luminance (and thus, the actual projected image). This shape is compared against the shape of the input feed in memory, and the relevant skew, rotation, etc. transformations are calculated, suc...