Digital image correlation (DIC), a powerful optical non-contact full-field deformation measurement technique, relies on the core principle of accurately tracking the grayscale changes of a speckle pattern on an object's surface before and after deformation. Therefore, acquiring high-quality, stable speckle images is fundamental to ensuring the accuracy and reliability of DIC measurements. However, complex lighting conditions in real-world testing environments often become a major factor interfering with image quality and introducing measurement errors. This paper will explore the specific mechanisms by which complex lighting affects DIC measurement accuracy and propose practical countermeasures to help users overcome the challenges posed by light and obtain more reliable data.
I. Illumination: A Double-Edged Sword in DIC Measurement
Ideally, DIC measurements require a uniform, stable, moderate, and interference-free lighting environment to ensure:
High-contrast speckle: Clearly distinguishes speckles from the background.
High signal-to-noise ratio (SNR): The effective information (speckle) in the image is much higher than the noise.
Gray value stability: The gray value distribution of speckle changes only due to the deformation of the object, and does not change due to the fluctuation of light.
Non-interference: No additional light spots, shadows or reflections interfere with the speckle pattern.
Complex lighting refers to any lighting environment that deviates from the ideal state mentioned above. It directly affects image quality and introduces systematic or random errors through various stages of the DIC algorithm (image acquisition, subpixel interpolation, and image matching calculation), ultimately leading to deviations or increased noise in displacement and strain measurement results.
II. The Influence Mechanism of Complex Illumination on DIC Accuracy
Insufficient lighting (low brightness):
Impact: Overall image grayscale values are low, speckle contrast with background is reduced, and image signal-to-noise ratio (SNR) is significantly decreased. Camera sensor readout noise and shot noise are relatively more pronounced.
Consequences: The DIC algorithm struggles to accurately identify and track speckle centers, leading to increased matching uncertainty and a significant increase in displacement/strain field noise (standard deviation), which may be completely overwhelmed by noise, especially in low-strain regions. Subpixel interpolation accuracy also decreases.
Excessive light intensity (overexposure/saturation):
Impact: When the grayscale value of a local or overall image reaches the saturation limit of the camera sensor (e.g., 255), speckle details are lost (becoming "white blocks"), and effective information is destroyed.
Consequences: Speckle information in saturated regions becomes completely invalid, making matching calculations impossible and resulting in missing local data. Even if not fully saturated, excessive illumination can introduce errors due to the sensor's nonlinear response. Halo artifacts are prone to occur at the edges of highlight areas.
Uneven lighting:
Impact: Inconsistent light intensity received by different areas of the surface of the object being measured results in bright and dark areas (halos or patches) in the image. The "apparent" grayscale of speckle depends not only on the speckle itself but also on the local light intensity.
as a result of:
Local contrast differences: Dark areas have poor contrast and high noise; bright areas may be saturated. Overall image quality is inconsistent.
Introducing spurious strain: When calculating displacement gradient (strain), the DIC algorithm misjudges the gray-level gradient caused by uneven illumination as the gray-level gradient caused by deformation, thus introducing a systematic spurious strain field. This is one of the most insidious and harmful effects.
Increased matching difficulty: The same speckle pattern has different appearance characteristics in different lighting areas, which increases the difficulty and error of cross-regional matching.
Unstable lighting (flickering/fluctuation):
Impact: Fluctuations in light source intensity over time (such as power frequency flicker caused by AC power supply, LED aging, and changes in ambient light) cause variations in the grayscale value of the same physical point across consecutive frames of the image.
Consequences: The DIC algorithm misinterprets the time changes of gray values as spatial displacement (i.e., object motion), thus introducing time-related noise or drift into the displacement and strain results, which seriously affects the accuracy of dynamic or long-term static measurements.
Directional strong light/shadow:
Impact: Point light sources or strong directional light sources (such as direct sunlight or unsoftened flash) will produce strong highlights (specular reflections) and shadow areas on the surface of objects with unevenness or structural features.
as a result of:
Highlights (glare): Cause severe overexposure and saturation in certain areas, loss of speckle information, and may produce large-area spot interference with adjacent areas.
Shadows: Cause insufficient local lighting, reduced contrast and SNR, and may even completely obscure speckle.
Altering the appearance of speckle: Strong shadow edges or highlight edges can “contaminate” the speckle pattern, introducing additional grayscale gradient interference.
Overall impact: Causes large areas of data to become unavailable and introduces significant matching errors and spurious strain in the light-dark boundary region.
Ambient light interference (stray light):
Impact: Uncontrollable external light sources (such as natural light from windows or other indoor lighting) enter the imaging system.
Consequences: When superimposed on target illumination, this can lead to unexpected changes in light intensity, uniformity, or spectral composition, collectively causing the aforementioned problems (such as intensity variations, non-uniformity, and instability). This presents a significant challenge, especially when conducting outdoor testing or in field tests where the light path cannot be completely sealed.
To obtain high-precision and reliable DIC measurement results, illumination control must be considered a core element of equal importance to system calibration and speckle preparation. By carefully designing active uniform illumination, strictly controlling the test environment, optimizing camera settings, ensuring high-quality speckle, and supplementing with appropriate image preprocessing and the selection of robust algorithms, the challenge of "light" can be overcome to the greatest extent.
