Anamorphic Speckle Patterns: Fixing DIC for Large Structures

Published in Optics and Lasers in Engineering (2026).

The Problem with DIC on Large Structures

Digital Image Correlation (DIC) is a standard tool for measuring deformations in experimental mechanics. It works by tracking a random speckle pattern on a surface between images. But when you’re measuring large structures like aircraft wings or fuselages, you often can’t position the camera perpendicular to the surface.

This creates a real problem: perspective distortions cause the speckles to appear non-uniform in the camera view, degrading correlation quality and increasing measurement uncertainty.

Figure 2: Perspective distortion effect on speckle patterns. Left: circular speckles viewed perpendicular to camera. Right: same speckles viewed from tilted angle — they appear deformed and non-uniform in size.

The Anamorphic Solution

The key insight is simple: if you know the camera geometry, you can pre-distort the speckle pattern so it appears uniform to the camera.

Think of it like anamorphic art — the image looks stretched when viewed from certain angles, but appears normal from a specific viewpoint. Here, the “anamorphic” speckle pattern is designed to compensate for the camera’s perspective, so the DIC algorithm sees uniform speckles.

Figure 3: Anamorphic transformation concept. Shows how a circular speckle is geometrically transformed so it appears uniform from the camera's viewpoint, even when projected onto a tilted surface.

How It Works

1. Geometric model — Known camera position and surface geometry define the perspective transformation
2. Pattern generation — A reference speckle pattern is locally transformed according to this transformation
3. Physical application — The anamorphic pattern is printed and applied to the test surface
4. Standard DIC — No changes to the correlation algorithm; it sees uniform speckles

Experimental Validation

The team tested this on a planar target with controlled tilt angles up to 89.7° — extreme grazing angles representative of real aircraft testing scenarios.

Figure 4: Experimental setup diagram. Reference plane with regular speckle (SP-Ref) at 2.5m from cameras. Tilted plane with both regular (SP-1) and anamorphic (SP-2) patterns for comparison.

Figure 5: Physical setup. Wooden boards on variable-height tripods allow controlled tilt angles. Stereo cameras mounted on fixed support, large LED panel provides uniform illumination.

Key results:

Figure 8: Main result — matching uncertainty across all tilt angles. The anamorphic pattern (SP-2, red) shows consistently lower uncertainty than the regular pattern (SP-1, blue). Mean values (bottom graph) confirm SP-2 outperforms SP-1 across all angles.

• Consistent uncertainty reduction — The anamorphic pattern consistently reduced matching uncertainty compared to regular patterns
• Improved 3D localization — Better precision in determining 3D positions of surface points
• Robust to angle changes — Benefits maintained even as camera incidence angles increased

Why This Matters

Aircraft testing often requires measuring large structures from fixed positions that create poor viewing angles. Traditional DIC setups struggle with these configurations. The anamorphic approach offers a practical solution:

Figure 11: 3D point cloud comparison. The anamorphic pattern (SP-2, red) produces points that cluster more tightly around the best-fit plane compared to the regular pattern (SP-1, blue), demonstrating improved 3D localization precision.

• No camera repositioning needed — The pattern compensates for the camera angle
• Works with existing DIC equipment — No special cameras or algorithms required
• Extends measurement range — Enables testing from positions that were previously too distorted for reliable DIC

Next Steps

While this study validates the approach on planar surfaces, the method is designed to extend to complex curved surfaces (aircraft wings, fuselages) and full-scale structures. The transformation framework is generic and can be applied offline during pattern generation.

---

The paper: “Planar validation of speckle pattern optimization for large-scale 3D streamlined structures in stereo digital image correlation using anamorphic transformation” by Stéphane W. Hu, Guilhem Marchal, Yvan Dilem, Denis Walch, Ilyass Tabiai. Optics and Lasers in Engineering, 202(2026) 109773. DOI.