SEMI-GLOBAL MATCHING (SGM) COST COMPRESSION

    公开(公告)号:US20220094933A1

    公开(公告)日:2022-03-24

    申请号:US17539945

    申请日:2021-12-01

    Abstract: This disclosure describes techniques for performing semi-global matching (SGM) path cost compression. In some examples, the techniques may perform disparity-dependent sub-sampling of a set of SGM path costs where the sub-sampling ratio is determined based on a candidate disparity level. The sub-sampled SGM path costs may be stored in a memory. When retrieved from memory, the sub-sampled SGM path costs may be interpolated to reconstruct the other path costs not stored in the memory. The reconstructed path costs may be used for further SGM processing. In further examples, the techniques may perform disparity-dependent quantization on the SGM path costs or the sub-sampled SGM path costs, and store the quantized SGM path costs in memory. The techniques of this disclosure may reduce bandwidth as well as reduce the memory footprint needed to implement an SGM algorithm.

    Semi-global matching (SGM) cost compression

    公开(公告)号:US11218699B2

    公开(公告)日:2022-01-04

    申请号:US16794551

    申请日:2020-02-19

    Abstract: This disclosure describes techniques for performing semi-global matching (SGM) path cost compression. In some examples, the techniques may perform disparity-dependent sub-sampling of a set of SGM path costs where the sub-sampling ratio is determined based on a candidate disparity level. The sub-sampled SGM path costs may be stored in a memory. When retrieved from memory, the sub-sampled SGM path costs may be interpolated to reconstruct the other path costs not stored in the memory. The reconstructed path costs may be used for further SGM processing. In further examples, the techniques may perform disparity-dependent quantization on the SGM path costs or the sub-sampled SGM path costs, and store the quantized SGM path costs in memory. The techniques of this disclosure may reduce bandwidth as well as reduce the memory footprint needed to implement an SGM algorithm.

    SEMI-GLOBAL MATCHING (SGM) COST COMPRESSION
    3.
    发明申请

    公开(公告)号:US20200186801A1

    公开(公告)日:2020-06-11

    申请号:US16794551

    申请日:2020-02-19

    Abstract: This disclosure describes techniques for performing semi-global matching (SGM) path cost compression. In some examples, the techniques may perform disparity-dependent sub-sampling of a set of SGM path costs where the sub-sampling ratio is determined based on a candidate disparity level. The sub-sampled SGM path costs may be stored in a memory. When retrieved from memory, the sub-sampled SGM path costs may be interpolated to reconstruct the other path costs not stored in the memory. The reconstructed path costs may be used for further SGM processing. In further examples, the techniques may perform disparity-dependent quantization on the SGM path costs or the sub-sampled SGM path costs, and store the quantized SGM path costs in memory. The techniques of this disclosure may reduce bandwidth as well as reduce the memory footprint needed to implement an SGM algorithm.

    Semi-global matching (SGM) cost compression

    公开(公告)号:US11831874B2

    公开(公告)日:2023-11-28

    申请号:US17539945

    申请日:2021-12-01

    Abstract: This disclosure describes techniques for performing semi-global matching (SGM) path cost compression. In some examples, the techniques may perform disparity-dependent sub-sampling of a set of SGM path costs where the sub-sampling ratio is determined based on a candidate disparity level. The sub-sampled SGM path costs may be stored in a memory. When retrieved from memory, the sub-sampled SGM path costs may be interpolated to reconstruct the other path costs not stored in the memory. The reconstructed path costs may be used for further SGM processing. In further examples, the techniques may perform disparity-dependent quantization on the SGM path costs or the sub-sampled SGM path costs, and store the quantized SGM path costs in memory. The techniques of this disclosure may reduce bandwidth as well as reduce the memory footprint needed to implement an SGM algorithm.

    Hybrid tiling strategy for semi-global matching stereo hardware acceleration

    公开(公告)号:US10080007B2

    公开(公告)日:2018-09-18

    申请号:US15073078

    申请日:2016-03-17

    Abstract: Apparatus and methods for disparity cost computation for a stereoscopic image. In an example, a method includes computing partial path matching costs for multiple disparities of pixels of multiple parallelogram blocks of an image in a forward pass direction. Between computing the partial path matching costs of the pixels of adjacent parallelogram blocks, the method includes computing partial path matching costs for the plurality of disparities of pixels along a boundary edge of a corresponding rectangular tile using paths that are external to the rectangular tile and computing partial path matching costs the disparities for all of the pixels of the corresponding rectangular tile. The method also includes computing total aggregated disparity costs for each disparity of the pixels of the image using the partial path matching costs computed for the pixels in the parallelogram blocks and in the rectangular tiles.

    SEMI-GLOBAL MATCHING (SMG) COST COMPRESSION
    6.
    发明申请
    SEMI-GLOBAL MATCHING (SMG) COST COMPRESSION 审中-公开
    SEMI全球匹配(SMG)成本压缩

    公开(公告)号:US20160227237A1

    公开(公告)日:2016-08-04

    申请号:US15012829

    申请日:2016-02-01

    Abstract: This disclosure describes techniques for performing semi-global matching (SGM) path cost compression. In some examples, the techniques may perform disparity-dependent sub-sampling of a set of SGM path costs where the sub-sampling ratio is determined based on a candidate disparity level. The sub-sampled SGM path costs may be stored in a memory. When retrieved from memory, the sub-sampled SGM path costs may be interpolated to reconstruct the other path costs not stored in the memory. The reconstructed path costs may be used for further SGM processing. In further examples, the techniques may perform disparity-dependent quantization on the SGM path costs or the sub-sampled SGM path costs, and store the quantized SGM path costs in memory. The techniques of this disclosure may reduce bandwidth as well as reduce the memory footprint needed to implement an SGM algorithm.

    Abstract translation: 本公开描述了用于执行半全局匹配(SGM)路径成本压缩的技术。 在一些示例中,这些技术可以对基于候选视差水平确定子采样比的一组SGM路径成本执行与视差相关的子采样。 子采样SGM路径开销可以存储在存储器中。 当从存储器检索时,子采样的SGM路径成本可被内插以重构未存储在存储器中的其他路径成本。 重建路径成本可用于进一步的SGM处理。 在另外的示例中,这些技术可以对SGM路径成本或子采样的SGM路径成本执行与视差相关的量化,并将量化的SGM路径成本存储在存储器中。 本公开的技术可以减少带宽以及减少实现SGM算法所需的存储器占用空间。

Patent Agency Ranking