Projection technology typically places several constraints on the geometric relationship between the projector and the projection surface to obtain an undistorted, properly sized image. In this paper we describe a simple, robust, fast, and low-cost method for automatic projector calibration that eliminates many of these constraints. We embed light sensors in the target surface, project Gray-coded binary patterns to discover the sensor locations, and then prewarp the image to accurately fit the physical features of the projection surface. This technique can be expanded to automatically stitch multiple projectors, calibrate onto nonplanar surfaces for object decoration, and provide a method for simple geometry acquisition. Categories and Subject Descriptors: H.5.1 [Multimedia Information Systems]: Augmented reality, I.3.6 [Methodology and Techniques]: Interaction techniques
Johnny C. Lee, Paul H. Dietz, Dan Maynes-Aminzade,