[sridhar-git05]

Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

• I agree to contribute to the project under Apache 2 License.
• To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
• The PR is proposed to the proper branch
• There is a reference to the original bug report and related work
• There is accuracy test, performance test and test data in opencv_extra repository, if applicable
  Patch to opencv_extra has the same branch name.
• The feature is well documented and sample code can be built with the project CMake

Fixes #4112.

StereoSGM stored the aggregated path costs for all paths in a single one-row buffer. The buffer size was calculated as:

width * height * numDisparities * num_paths

For sufficiently large images and disparity counts, this value can exceed the range of the signed 32-bit integer used by the GpuMat allocation interface.

Store each path in a separate row instead. The resulting layout is:

rows = num_paths
cols = width * height * numDisparities

This keeps the per-row size within the existing integer-based GpuMat interface while preserving the same logical path/disparity data.

Updated PathAggregation and WinnerTakesAll to use the new row-based layout and updated the CPU reference test accordingly.

Added a high-level StereoSGM regression test covering MODE_HH with 256 disparities.

Testing:

• StereoSGM_AggregatedBufferLayout: 1/1 passed
• StereoSGM_WinnerTakesAll: 8/8 passed
• StereoSGM_PathAggregation: 96/96 passed
• Total focused tests: 105/105 passed
• git diff --check passes