Use of Projector-Camera System for Human-Computer Interaction.

摘要

The use of a projector in place of traditional display device would dissociate display size from device size, making portability much less an issue. Associated with camera, the projector-camera system allows simultaneous video display and 3D acquisition through imperceptible structured light sensing, providing a vivid and immersed platform for natural human-computer interaction. Key issues involved in the approach include: (1) Simultaneous Display and Acquisition: how to make normal video projector not only a display device but also a 3D sensor even with the prerequisite of incurring minimum disturbance to the original projection; (2) 3D Information Interpretation: how to interpret the spare depth information with the assistance of some additional cues to enhance the system performance; (3) Segmentation: how to acquire accurate segmentation in the presence of the incessant variation of the projected video content; (4) Posture Recognition: how to infer 3D posture from single image. This thesis aims at providing improved solutions to each of these issues.;To address the conflict between imperceptibility of the embedded codes and the robustness of code retrieval, noise-tolerant schemes to both the coding and decoding stages are introduced. At the coding end, specifically designed primitive shapes and large Hamming distance are employed to enhance tolerance toward noise. At the decoding end, pre-trained primitive shape detectors are used to detect and identify the embedded codes -- a task difficult to achieve by segmentation that is used in general structured light methods, for the weakly embedded information is generally interfered by substantial noise.;On 3D information interpretation, a system that estimates 6-DOF head pose by imperceptible structured light sensing is proposed. First, through elaborate pattern projection strategy and camera-projector synchronization, pattern-illuminated images and the corresponding scene-texture image are captured with imperceptible patterned illumination. Then, 3D positions of the key facial feature points are derived by a combination of the 2D facial feature points in the scene-texture image localized by AAM and the point cloud generated by structured light sensing. Eventually, the head orientation and translation are estimated by SVD of a correlation matrix that is generated from the 3D corresponding feature point pairs over different frames.;On the segmentation issue, we describe a coarse-to-fine hand segmentation method for projector-camera system. After rough segmentation by contrast saliency detection and mean shift-based discontinuity-preserved smoothing, the refined result is confirmed through confidence evaluation.;Finally, we address how an HCI (Human-Computer Interface) with small device size, large display, and touch input facility can be made possible by a mere projector and camera. The realization is through the use of a properly embedded structured light sensing scheme that enables a regular light-colored table surface to serve the dual roles of both a projection screen and a touch-sensitive display surface. A random binary pattern is employed to code structured light in pixel accuracy, which is embedded into the regular projection display in a way that the user perceives only regular display but not the structured pattern hidden in the display. With the projection display on the table surface being imaged by a camera, the observed image data, plus the known projection content, can work together to probe the 3D world immediately above the table surface, like deciding if there is a finger present and if the finger touches the table surface, and if so at what position on the table surface the finger tip makes the contact. All the decisions hinge upon a careful calibration of the projector-camera-table surface system, intelligent segmentation of the hand in the image data, and exploitation of the homography mapping existing between the projector's display panel and the camera's image plane.