Overview
ISO 16610-29:2020 - "Geometrical product specifications (GPS) - Filtration - Part 29: Linear profile filters: wavelets" defines the terminology, concepts and requirements for using biorthogonal wavelets as linear profile filters in surface metrology. The standard specifies compact‑support biorthogonal wavelets for profile decomposition, explains the discrete wavelet transform (DWT) and multiresolution analysis, and gives normative procedures for two wavelet families (cubic prediction and cubic B‑spline) in annexes.
Key topics and requirements
- Scope and terminology: Basic GPS vocabulary for wavelets (mother wavelet, scaling function, wavelet function, multiresolution, decimation, lifting scheme).
- Biorthogonal wavelets: Specification of invertible (not necessarily orthogonal) wavelet transforms that allow symmetric functions and linear‑phase filtering.
- Discrete wavelet transform (DWT): Principles of decomposing uniformly sampled profile data into low‑pass (smooth) and high‑pass (detail) components with decimation (typically down‑sampling by 2).
- Multiresolution analysis: Ladder structure of successive smoothing and detail levels enabling lossless reconstruction of the original profile.
- Lifting scheme: Fast wavelet implementation using splitting, prediction and updating stages.
- Filter coefficients & designation: Defining low‑pass (scaling) and high‑pass (wavelet) filter coefficient sets used in analysis.
- Normative annexes:
- Annex A - requirements/procedure for cubic prediction wavelets.
- Annex B - requirements/procedure for cubic B‑spline wavelets.
- Informative annexes: Relationships to the filtration matrix model and the wider GPS matrix model (for traceability and integration with other filtering frameworks).
Applications and users
ISO 16610-29:2020 is intended for professionals involved in surface texture measurement, filtering and analysis:
- Metrology engineers and laboratory staff performing profile filtering and surface characterization.
- Quality‑control and manufacturing engineers analyzing localized defects (scratches, irregularities) and non‑stationary profiles.
- Software developers and instrument manufacturers implementing wavelet‑based filtering and DWT algorithms for profilometers and surface analysis tools.
- Standards committees and technical authors aligning GPS filtration methods across the ISO 16610 series.
Wavelet filters are particularly useful when locating features in both space and scale (e.g., localized defects), achieving linear‑phase filtering with symmetric wavelets, and performing multiscale noise separation and feature extraction.
Related standards
Keywords: ISO 16610-29:2020, GPS filtration, biorthogonal wavelets, linear profile filters, discrete wavelet transform, multiresolution, cubic B‑spline, cubic prediction, surface texture metrology.