By treating projectors as pin-hole cameras, we show it is possible to calibrate the projectors of a casually-aligned, multi-projector display wall using the principles of planar auto-calibration. We also use a pose estimation technique for planar scenes to reconstruct the relative pose of a calibration camera, the projectors and the plane they project on. Together with assumptions about the pose of the camera, we use the reconstruction to automatically compute the projector-display homographies needed to render properly scaled and oriented imagery on the display wall. The main contribution of this paper is thus to provide a fully automated approach to calibrate a multi-projector display wall without the need for fiducials or interaction.

Arguably, the most vexing problem remaining for multi-projector displays is that of photometric (brightness) seamlessness within and across different projectors. Researchers have strived for strict photometric uniformity that achieves identical response at every pixel of the display. However, this goal typically results in displays with severely compressed dynamic range and poor image quality. In this paper, we show that strict photometric uniformity is not a requirement for achieving photometric seamlessness. We introduce a general goal for photometric seamlessness by defining it as an optimization problem balancing perceptual uniformity with display quality. Based on this goal, we present a new method to achieve perceptually seamless high quality displays. We first derive a model that describes the photometric response of projection-based displays. Then we estimate the model parameters and modify them using perception-driven criteria. Finally, we use the graphics hardware to reproject the image computed using the modified model parameters by manipulating only the projector inputs at interactive rates. Our method has been successfully demonstrated on three different practical display systems at Argonne National Laboratory, made of 2 × 2 array of four projectors, 2 × 3 array of six projectors and 3 × 5 array of fifteen projectors. Our approach is efficient, automatic and scalable – requiring only a digital camera and a photometer. To the best of our knowledge, this is the first approach and system that addresses the photometric variation problem from a perceptual stand point and generates truly seamless displays with high dynamic range.

A theory for photometric calibration of cameras and multiple projectors with overlapping displays is presented. The theory is predominantly based on the analysis of isointensity curves in projector input state space – curves which define different projector input intensity combinations that result in the same camera-observed pixel intensities. Three methods, which have different speed and accuracy tradeoffs, are proposed for recovering the projector-to-screen and screento-camera intensity transfer functions. The methods do not require a specific parametric model for the shapes of these functions, nor impose any smoothness constraints. Additional methods are described for calibrating projector offsets and binary light sources, and also extending the perpixel analysis to other pixels in the display. The methods do not require the use of expensive equipment, and may be carried out with a low dynamic range camera with minimal controls.

Figure 1: A two-projector display after both projectors have been moved (left) and after roughly ten seconds of operation (middle,right). We present a novel calibration framework for multi-projector displays that achieves continuous geometric calibration by estimating and refining the poses of all projectors in an ongoing fashion during actual display use. Our framework provides scalability by operating as a distributed system of “intelligent ” projector units: projectors augmented with rigidly-mounted cameras, and paired with dedicated computers. Each unit interacts asynchronously with its peers, leveraging their combined computational power to cooperatively estimate the poses of all of the projectors. In cases where the projection surface is static, our system is able to continuously refine all of the projector poses, even when they change simultaneously.

Fig. 1. Left: A 2 × 4 array of eight projectors on a cylindrical display surface showing a weather map visualization; Right: A 2 × 3 array of six projectors on a more general extruded surface showing a medical visualization. Abstract—In this paper, we present the first algorithm to geometrically register multiple projectors in a view-independent manner (i.e. wallpapered) on a common type of curved surface, vertically extruded surface, using an uncalibrated camera without attaching any obtrusive markers to the display screen. Further, it can also tolerate large non-linear geometric distortions in the projectors as is common when mounting short throw lenses to allow a compact set-up. Our registration achieves sub-pixel accuracy on a large number of different vertically extruded surfaces and the image correction to achieve this registration can be run in real time on the GPU. This simple markerless registration has the potential to have a large impact on easy set-up and maintenance of large curved multi-projector displays, common for visualization, edutainment, training and simulation applications. Index Terms—Registration, Calibration, Multi-Projector Displays, Tiled Displays. 1

In this paper we present a novel technique to calibrate multiple casually aligned projectors on a fiducial-free cylindrical curved surface using a single camera. We impose two priors to the cylindrical display: (a) cylinder is a vertically extruded surface; and (b) the aspect ratio of the rectangle formed by the four corners of the screen is known. Using these priors, we can estimate the display’s 3D surface geometry and camera extrinsic parameters using a single image without any explicit display to camera correspondences. Using the estimated camera and display properties, we design a novel deterministic algorithm to recover the intrinsic and extrinsic parameters of each projector using a single projected pattern seen by the camera which is then used to register the images on the display from any arbitrary viewpoint making it appropriate for virtual reality systems. Finally, our method can be extended easily to handle sharp corners- making it suitable for the common CAVE like VR setup. To the best of our knowledge, this is the first method that can achieve accurate geometric auto-calibration of multiple projectors on a cylindrical display without performing an extensive stereo reconstruction.

a) b) Figure 1: Foveal inset. a: The user directs the projection of the foveal inset using a laser pointer. The size of the projected inset is significantly smaller than the tiles of the rear-projected display wall, thus providing a higher resolution. b: Magnified view of a high-resolution slice of a cryosection of a monkey brain. The boundary between the high-resolution foveal inset (left) and the lower-resolution display wall (right) is clearly visible. Note that the pixel dimensions of the foveal inset projector and the display wall projector are identical. (Aerial photographs courtesy of the City of Davis, CA. Monkey brain data set courtesy of E.G. Jones, UCD Center for Neuroscience.) We introduce a system that adds a foveal inset to large-scale projection displays. The effective resolution of the foveal inset projection is higher than the original display resolution, allowing the user to see more details and finer features in large data sets. The foveal inset is generated by projecting a high-resolution image onto a mirror mounted on a pan–tilt unit (PTU) that is controlled by the user with a laser pointer. Our implementation is based on Chromium and supports many OpenGL applications without modifications.

Display and interaction real-estate is fundamentally limited by the size of the screen in traditional display devices. Projection-based interfaces allow interaction with appropriated, everyday, passive surfaces. Instead of restricting interaction to a single rectangle, projection-based augmented reality makes interaction ubiquitous, taking the computer out of the box. Extending interaction to almost any surface enables new possibilities in a host of new domains, including entertainment, consumer electronics, advertising, medicine and many more. Overlaying virtual elements onto complex physical surfaces under real-world conditions, presents a variety of challenges, many which are open research problems. A wealth of previous work has focused on projection onto constrained surfaces like planes, parabolic curves, etc. In this research, we explore projection onto arbitrarily complex physical surfaces. We describe a complete, practical approach embodied in a projector-camera system. Our system is constructed of low-cost, commodity hardware and demonstrates some of the exciting applications of Spatial Augmented Reality. We present a novel initial homography estimation for interactive projector-camera calibration. We introduce a unique hybrid projector pattern codification technique for improved correspondences. Finally, we present interactive surface particles, which are a surface independent content representation that enables reusable interaction

We introduce a system that adds a foveal inset to large-scale projection displays. The effective resolution of the foveal inset projection is higher than the original display resolution, allowing the user to see more details and finer features in large data sets. The foveal inset is generated by projecting a high-resolution image onto a mirror mounted on a pan–tilt unit that is controlled by the user with a laser pointer. Our implementation is based on Chromium and supports many OpenGL applications without modifications. We present experimental results using high-resolution image data from medical imaging and aerial photography.

Multi-projector displays require a variety of image correction techniques. Geometric distortions resulting from projecting onto non-planar display surfaces and non-linear properties of projector lenses must be corrected to produce usable images. Photometric issues such as varying luminance response between projectors and brightness discontinuities in projector overlap regions must also be corrected in a high-quality display. We show how the programmability of modern graphics hardware can be used to e#ciently implement several image correction techniques for multi-projector displays.