Digital holography is a powerful approach for remote metrology at different scales (micro and macro). Holographic interferometry provides the optical path difference in a wrapped modulo 2pi phase. The phase connected to the scene/object/structure of interest can be helpfully considered for many industrial purposes: roughness measurements, surface shape profiling, surface deformation or vibration measurements. Holographic interferometry has the benefit of being contactless and provides full-field measurements. In addition, the use of light illumination makes it non-intrusive. High temporal resolution can be obtained when using high-speed cameras. However, the phase data from holographic interferometry suffers from speckle decorrelation which adds noise in the measurement.
In this webinar hosted by the Holography and Diffractive Optics Technical Group, Pascal Picart will explain the origin of this noise and draw the basic of theory. This will lead to highlight the very particular properties of the speckle noise in holographic metrology: non-Gaussian, non-stationary, amplitude-dependent (dependent on the local fringe density) and, according to the symmetry of the holographic system, possibly anisotropic. Then, ways to efficiently process the data will be given. Picart will detail approaches based on deep learning and discuss their perspectives for data processing in holographic metrology.
What You Will Learn:
• Fundamentals of digital holography and digital holographic interferometry
• Contactless metrology with holography
• Noise in measurements
• The algorithms for processing with special attention to the recent deep learning approaches
Who Should Attend:
• Researchers from the fields of coherent imaging and optical metrology