Abstract: Approaches are described for generating a multiview autostereoscopic image from a stereo three-dimensional input image pair. A stereo to multiview rendering system receives a stereo three-dimensional image including a left image and a right image at a first view position and a second view position, respectively. The system generates a first input warp and a second input warp that maps the left image and the right image to a third and fourth positions, respectively, where the third and fourth positions lie between the first and second view positions. The system generates a plurality of output warps based on the first warp and the second warp. The system resamples each output warp in the plurality of output warps to create a plurality of partial output images. The system interleaves the plurality of partial output images to generate a composite output image.

Abstract: A closed-loop control system for stereoscopic video capture is provided. At least two motorized lenses are positioned in accordance with specified parameters to capture spatially-disparate images of a scene. The motorized lenses focus light on a corresponding one of the at least two sensors, which generate image streams. One or more processors execute instructions to provide a stream analyzer and a control module. The stream analyzer receives the image streams from the sensors and analyzes the image streams and the specified parameters in real time; the stream analyzer then modifies the image streams and generates metadata. The control module then receives and analyzes the image streams and metadata and transmits updated parameters to a control mechanism that is coupled to the at least two motorized lenses. The control mechanism then modifies operation of the at least two motorized lenses in real time in accordance with the updated parameters.

Abstract: Techniques are disclosed for rendering images. The techniques include receiving an input image associated with a source space, the input image comprising a plurality of source pixels, and applying an adaptive transformation to a source pixel, where the adaptive transformation maps the source pixel to a target space associated with an output image comprising a plurality of target pixels. The techniques further include determining a target pixel affected by the source pixel based on the adaptive transformation. The techniques further include writing the transformed source pixel into a location in the output image associated with the target pixel.

Abstract: Techniques are disclosed for retargeting images. The techniques include receiving one or more input images, computing a two-dimensional saliency map based on the input images in order to determine one or more visually important features associated with the input images, projecting the saliency map horizontally and vertically to create at least one of a horizontal and vertical saliency profile, and scaling at least one of the horizontal and vertical saliency profiles. The techniques further include creating an output image based on the scaled saliency profiles. Low saliency areas are scaled non-uniformly while high saliency areas are scaled uniformly. Temporal stability is achieved by filtering the horizontal resampling pattern and the vertical resampling pattern over time. Image retargeting is achieved with greater efficiency and lower compute power, resulting in a retargeting architecture that may be implemented in a circuit suitable for mobile applications such as mobile phones and tablet computers.

Abstract: Approaches are described for generating a multiview autostereoscopic image from a stereo three-dimensional input image pair. A stereo to multiview rendering system receives a stereo three-dimensional image including a left image and a right image at a first view position and a second view position, respectively. The system generates a first input warp and a second input warp that maps the left image and the right image to a third and fourth positions, respectively, where the third and fourth positions lie between the first and second view positions. The system generates a plurality of output warps based on the first warp and the second warp. The system resamples each output warp in the plurality of output warps to create a plurality of partial output images. The system interleaves the plurality of partial output images to generate a composite output image.

Abstract: Techniques are disclosed for rendering images. The techniques include receiving an input image associated with a source space, the input image comprising a plurality of source pixels, and applying an adaptive transformation to a source pixel, where the adaptive transformation maps the source pixel to a target space associated with an output image comprising a plurality of target pixels. The techniques further include determining a target pixel affected by the source pixel based on the adaptive transformation. The techniques further include writing the transformed source pixel into a location in the output image associated with the target pixel.

Abstract: Techniques are disclosed for retargeting images. The techniques include receiving one or more input images, computing a two-dimensional saliency map based on the input images in order to determine one or more visually important features associated with the input images, projecting the saliency map horizontally and vertically to create at least one of a horizontal and vertical saliency profile, and scaling at least one of the horizontal and vertical saliency profiles. The techniques further include creating an output image based on the scaled saliency profiles. Low saliency areas are scaled non-uniformly while high saliency areas are scaled uniformly. Temporal stability is achieved by filtering the horizontal resampling pattern and the vertical resampling pattern over time. Image retargeting is achieved with greater efficiency and lower compute power, resulting in a retargeting architecture that may be implemented in a circuit suitable for mobile applications such as mobile phones and tablet computers.

Abstract: A closed-loop control system for stereoscopic video capture is provided. At least two motorized lenses are positioned in accordance with specified parameters to capture spatially-disparate images of a scene. The motorized lenses focus light on a corresponding one of the at least two sensors, which generate image streams. One or more processors execute instructions to provide a stream analyzer and a control module. The stream analyzer receives the image streams from the sensors and analyzes the image streams and the specified parameters in real time; the stream analyzer then modifies the image streams and generates metadata. The control module then receives and analyzes the image streams and metadata and transmits updated parameters to a control mechanism that is coupled to the at least two motorized lenses. The control mechanism then modifies operation of the at least two motorized lenses in real time in accordance with the updated parameters.