Scholarlink

Background <p>Digital Image Correlation (DIC) is a non-contact optical measurement method widely employed for measuring object motion and deformation. However, traditional 2D-DIC fails in scenarios involving large-angle specimen rotation or off-axis due to geometric correspondence violations and severe decorrelation.</p> Objective <p>To overcome this fundamental limitation, we introduce the Affine-Projection Transformation DIC (APT-DIC) method.</p> Methods <p>Our novel approach features a two-stage rectification mechanism driven by Speeded-Up Robust Features (SURF): an initial affine transformation rectifies in-plane rotational distortions, which is then seamlessly followed by a projective transformation to compensate for out-of-plane rotation effects. Rigorous validation experiments confirm the stable DIC computation capability of the method under geometric correspondence alterations and large rotations. Controlled variable tests demonstrate the Pareto optimality of APT in rectification efficiency while maintaining robustness within calibration regions containing irregular deformation fields.</p> Results <p>When applied to gear deformation monitoring, APT-DIC achieves remarkable agreement between the calculated full-field displacements, strain evolution trends, and reference values.</p> Conclusion <p>This work establishes a new paradigm for high-fidelity deformation characterization in rotating machinery and other scenarios involving large rotations, paving the way for reliable measurements in previously challenging environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Digital Image Correlation Under Large Rotations or Off-Axis: An APT-DIC Method with SURF-Driven Affine-Projective

  • 期刊论文

  • 发表日期：

    2026年4月20日

查看全文
  • X. Y. Hou,
  • D. W. Zeng,
  • H. Y. Zhang,
  • Z. W. Liu,
  • X. F. Huang

摘要

Background

Digital Image Correlation (DIC) is a non-contact optical measurement method widely employed for measuring object motion and deformation. However, traditional 2D-DIC fails in scenarios involving large-angle specimen rotation or off-axis due to geometric correspondence violations and severe decorrelation.

Objective

To overcome this fundamental limitation, we introduce the Affine-Projection Transformation DIC (APT-DIC) method.

Methods

Our novel approach features a two-stage rectification mechanism driven by Speeded-Up Robust Features (SURF): an initial affine transformation rectifies in-plane rotational distortions, which is then seamlessly followed by a projective transformation to compensate for out-of-plane rotation effects. Rigorous validation experiments confirm the stable DIC computation capability of the method under geometric correspondence alterations and large rotations. Controlled variable tests demonstrate the Pareto optimality of APT in rectification efficiency while maintaining robustness within calibration regions containing irregular deformation fields.

Results

When applied to gear deformation monitoring, APT-DIC achieves remarkable agreement between the calculated full-field displacements, strain evolution trends, and reference values.

Conclusion

This work establishes a new paradigm for high-fidelity deformation characterization in rotating machinery and other scenarios involving large rotations, paving the way for reliable measurements in previously challenging environments.