Multi-resolution depth estimation using modified census transform for advanced driver assistance systems
AssigneeIllinois Institute of Technology
InventorsMaziar Loghman, Maral Mesmakhosroshahi, Joohee Kim
Life of this patent
Estimating the term…
Abstract
A computer-implemented depth estimation method based on non-parametric Census transform with adaptive window patterns and semi-global optimization. A modified cross-based cost aggregation technique adaptively creates the shape of the cross for each pixel distinctly. In addition, a depth refinement algorithm fills holes within the estimated depth map using the surrounding background depth pixels and sharpens the object boundaries by exerting a trilateral filter to the generated depth map. The trilateral filter uses the curvature of pixels as well as texture and depth information to sharpen the edges.
Claims20
A method of determining depth in images, comprising:
- generating an image mask indicating smooth areas and object edges for each of a pair of stereo images;
- distinguishing between the smooth areas and the object edges as a function of pixel curvature;
- aggregating curvature values over a predetermined window;
- generating a curvature map for the each of a pair of stereo images;
- and
- generating the image mask from the curvature map;
- calculating a computation cost for a plurality of pixels within the image mask;
- and
- matching pixels between the each of a pair of stereo images and optimizing a value for the matching pixels to obtain an estimated depth map.
A method of determining depth in images, comprising:
- generating an image mask indicating smooth areas and object edges for each of a pair of stereo images;
- calculating a computation cost for a plurality of pixels within the image mask;
- aggregating a cost of each of the pixels over a support region within the pair of stereo images;
- creating an adaptive cross for each of the pixels;
- determining a horizontal and vertical arm length for the adaptive cross;
- aggregating cost values within the created adaptive cross;
- and
- matching pixels between the each of a pair of stereo images and optimizing a value for the matching pixels to obtain an estimated depth map.
A method of determining depth in images, comprising:
- reducing image resolution in each of a pair of stereo images;
- creating an image mask as a function of pixel curvatures for each of the pair of stereo images;
- determining a computational cost value for a plurality of pixels using a Census transform to map a local neighborhood surrounding each of the plurality of pixels to a bit stream;
- aggregating cost values for each of the plurality of pixels;
- and
- matching pixels between the pair of stereo images and optimizing a value for the matching pixels to obtain an estimated depth map;
- wherein the aggregating cost values comprises:
- creating an adaptive cross shape;
- and
- aggregating cost values within the created adaptive cross shape.
Dependencies are inferred from the claim text.