Abstract: Panoramic and spherical cameras often currently capture respective portions of a scene using a set of lenses that focus images on imagers for sampling by photosensitive elements. Because the images are often circular and imagers are often rectangular, the orientation of the lenses may be selected as a tradeoff between capturing capture the entire image while leaving unused portions of the imager, and maximizing imager coverage to increase resolution while creating cropped image portions that extend beyond the edges of the imager. The techniques presented herein involve displacing the image from the imager center in order to increase a first cropped area and reduce or eliminate a second cropped area. The lenses may also be oriented such that the first cropped area comprises a portion of the composite image that may be acceptable to omit from the composite image (e.g., the user's hand or a camera mounting apparatus).

Abstract: Panoramic and spherical cameras often currently capture respective portions of a scene using a set of lenses that focus images on imagers for sampling by photosensitive elements. Because the images are often circular and imagers are often rectangular, the orientation of the lenses may be selected as a tradeoff between capturing capture the entire image while leaving unused portions of the imager, and maximizing imager coverage to increase resolution while creating cropped image portions that extend beyond the edges of the imager. The techniques presented herein involve displacing the image from the imager center in order to increase a first cropped area and reduce or eliminate a second cropped area. The lenses may also be oriented such that the first cropped area comprises a portion of the composite image that may be acceptable to omit from the composite image (e.g., the user's hand or a camera mounting apparatus).

Abstract: Panoramic and spherical cameras are often configured to capture respective portions of a scene using a set of lenses that focus images on imagers for sampling by photosensitive elements. In many such cameras, the orientation of the lenses is selected to resemble to a regular prismatic solid (e.g., one lens oriented according to a face of a cube). However, such lens orientations may create gaps between images that result in blind spots, and/or varying degrees of coverage overlap. Presented herein are techniques for orienting the lenses in an asymmetric manner, comprising one forward lens and three backward lenses having a 120-degree rotational angle around a first (e.g., front-to-back) axis and a variable inclination angle perpendicular to the first axis. This lens orientation may be selected (e.g., by a computer) to achieve a desired degree of coverage overlap while significantly reducing gaps that create blind spots in the composite image.