SBIR-STTR Award

This Small Business Innovation Research Phase I project proposes to investigate a new imaging technique which involves dynamic control of the tilt of the camera sensor plane. The proposed effort will investigate algorithms using this geometry with two objectives. The first objective is to acquire the range map of a scene. This is useful to construct geometric models of objects, which are in turn useful for navigation, recognition and manipulation. Second, the imaging technique delivers a focused photograph of multiple objects. This is useful for photography because a larger part of the scene can be included in the foreground than possible with normal cameras. Traditionally, range sensing requires the use of an energy scan, such as using laser and range measurement from returned energy. Both of these methods involve special purpose illumination of the scene. The proposed work is aimed at fast acquisition of geometric models in natural lighting. Imaging of surfaces that have large variations in depth are not traditionally possible, so the proposed work has the potential of introducing new photographic capabilities. The proposed imaging geometry and algorithms exploit optics of image formation in place of mechanical control to achieve both objectives simultaneously

This Small Business Innovation Research Phase II project will develop a new imaging technology called nonfrontal imaging. The core of this new approach is active control of imaging geometry to eliminate the mechanical adjustment of sensor plane location usually required for focusing, and exploit imaging geometry and optics to achieve focusing from camera's panning motion alone. This not only results in high imaging speeds but also in seamless panoramic images. A nonfrontal imaging camera can obtain a panoramic focused image of all objects in an arbitrarily large scene regardless of their locations, which is in registration with a range map acquired in parallel. Development of algorithms and the building of a working prototype are proposed which will have the following capabilities: panoramic focused image acquisition, panoramic range image acquisition, and omni-focused 3D stereo display. All these capabilities will be achieved at multiple resolutions and for up to 360-degree views of the scene. Sensor surface orientation will be controlled in a dynamic, data-driven mode. The prototype will be tested primarily in applications involving photography and surveillance, and to a lesser extent surgery and manufacturing. The system development will be done in consultation with industry. It is expected that the resulting technology will introduce hitherto unavailable capabilities as well as lead to significant strides in existing capabilities. A US patent on the camera has been awarded, and foreign patents have been filed. Potential commercial applications are: Wide-scene studio photography; Outdoor nature photography; Endoscopic and neurosurgery; Television broadcasting; Geometric modeling for computer aided manufacturing; Surveillance of homes, stores and compounds; Virtual reality using omni-focused 3D display; Advertising; Visual art; Monitoring and control of hazardous environments; Visualization; Interactive video games