16 September 2025 in Measurement
31 December 2025 in IEEE Sensors Journal
To solve the problem of high mismatching rate in traditional three-dimensional (3D) stereo matching algorithms in uneven light and weak textures, an enhanced cross-based cost aggregation semi-global block matching (ECCA-SGBM) algorithm is proposed for the 3D reconstruction of aluminum alloy casting with weak textures and reflective surfaces. First, an experimental binocular-vision platform is constructed. Subsequently, a fusion cost calculation method based on Scharr operator and an enhanced adaptive cross-based cost aggregation method are introduced to compute disparity and depth map. Finally, the depth map is converted into point cloud data, which is used to reconstruct the 3D structure of aluminum alloy casting. Experimental results demonstrate that the proposed ECCA-SGBM matching algorithm reduces the mismatching rate in the areas with uneven light effectively. The average error in the aperture diameter of the reconstructed 3D point cloud of the aluminum alloy casting is less than 5%, which can serve as a useful reference for the rapid size measurement of other machine parts.
24 August 2026 in Measurement
23 January 2026 in Measurement
15 May 2026 in Measurement
9 July 2026 in Measurement