The novelty in a SLAM claim is usually the choice of what to track. Huawei's grant US11270148B2 ("Visual SLAM method and apparatus based on point and line features," issued March 8, 2022; inventors Wanlong Li, Jianfei Li, Yajun Gao) commits to both: point features and line features, used together. The combination is the claim, and it's a direct answer to a known weakness of point-only SLAM.

The mechanism extracts and tracks two primitive types simultaneously. Point features (corners, blobs) are abundant but fragile in low-texture scenes; line features (edges, structural boundaries) persist where points vanish — along blank walls, hallways, building facades. Fusing both gives a map that stays observable in environments that would starve a point-only tracker. G06K 9/4609 (feature extraction) and 9/00664 (3D scene recognition) mark the perception machinery.

“Systems and methods of visual simultaneous localization and mapping (SLAM) based on point and line features are disclosed.”— U.S. Patent No. 11,270,148 source

For the control beat, the engineering logic is sound and the IP logic follows it: by fencing the point-and-line combination, the claim covers the robustness improvement that makes visual SLAM work in the structured, low-texture indoor and urban scenes where it otherwise drifts. It's a narrow conceptual delta over point-only SLAM, but a practically important one — and a clean place to carve patent space in a crowded field.

From a portfolio angle, Huawei files heavily across perception and computer vision, and a point-line SLAM grant is a building block in a broad spatial-computing portfolio that spans phones, AR, and autonomy. The strategic read is less about this single claim and more about Huawei's accumulating density in visual localization — a portfolio that becomes a cross-licensing chip as spatial computing matures.

Caveats. Point-line SLAM has explicit academic precedent (PL-SLAM and related work), so the grant turns on the specific extraction-and-fusion method in claim 1, not on the idea of combining primitives. The G06K codes mark feature-based vision; they don't define the algorithm. Read the independent claim for how lines and points are jointly optimized.

For the file: a robustness-motivated point-line visual-SLAM grant inside a broad Huawei vision portfolio. Pull US11270148B2 in the patent record, read claim 1 for the joint point-line optimization step, and treat it as one tile in Huawei's spatial-localization mosaic.