⚡
Accelerated Time-to-Deployment
A validation pilot typically runs 2-4 weeks. Integration is a standard REST API with an OpenAPI specification.
Artifact-backed deployment methodology
Deterministic Physics — Not Machine Learning
Software-Defined
Precision Metrology
Physics-based holographic reconstruction platform producing quantitative phase and height maps from standard industrial imaging sensors, with edge-aware filtering and SNR-adaptive fusion. No AI training required — the same input and config produce the same result, and anyone can recompute it.
Physics-based (no ML)
Paid pilots now open
Deterministic & recomputable
Core Technical Capabilities
Holographic reconstruction with a determinism contract and provenance on every result.
✓
Zero-Phase Filtering
Forward-backward filtering eliminates spatial drift in fringe patterns.
Spatial Drift
<0.05 px
★
Edge-Aware Enhancement
Bilateral filtering preserves sharp transient details in reconstructions.
Edge Preservation
Edge-aware
✓
Block Overlap-Add
Perfect reconstruction with seamless block merging and windowing.
Reconstruction MSE
<1e-15
★
SNR-Adaptive Fusion
Intelligent regime-based weighting for dual-path processing.
Automatic regime detection: HIGH/MID/LOW
RESTful API Integration
Enterprise-grade API designed for production MES and QA workflows. Standard HTTP endpoints with comprehensive OpenAPI documentation.
Integration Example
POST https://api.apertureapi.com/reconstruct_pdhm
?wavelength_um=0.532
&pixel_pitch_um=3.45
&z_min_mm=0.5&z_max_mm=3.0
&quality_mode=max_snr
Headers: Authorization: Bearer {api_key}
Content-Type: multipart/form-data
Body:
file: [TIFF/PNG/NPY/JPG]
Optical parameters are query parameters.
Response: a ZIP archive.
Accepts standard imaging formats. Returns quantitative phase and height maps as 3D reconstructions in GLB, OBJ, STL, or PLY format, with quality metrics and enhancement data.
Deterministic
Reconstruction
Physics-based computational optics — no machine learning uncertainty.
No Training Required: Deterministic algorithms improve through volume processing practice, not ML training.
A second measurement point: an independent, recomputable opinion on data you already have — beside your existing instrument, not instead of it.
Advanced: edge-aware filtering with intelligent SNR-adaptive fusion.
A Second Measurement Point, Without Capex
Advanced platform delivering enterprise metrology without hardware acquisition. Delivering enterprise metrology capabilities.
Zero Hardware Lock-In
Deploy with Existing Infrastructure
Compatible with industrial imaging sensors and standard coherent light sources. No vendor lock-in or proprietary hardware required.
Physics-Based Certainty
No AI Training Required
Computational optics based on wave propagation physics. Deterministic, auditable results — no machine learning uncertainty or training datasets.
Recomputable Results
Recomputable Results
Software-defined metrology with a determinism contract: pinned seeds and BLAS threads, a content-addressed report ID that rehashes, and SHA-256 commitments to input and config. Edge-aware filtering and SNR-adaptive fusion throughout.
Fast Time-to-Value
2-4 Week Integration
Standard deployment timeline from evaluation to production. Platform designed for seamless integration.
Platform Architecture: Physics-Based Processing Pipeline
Advanced deterministic holographic reconstruction through GPU-accelerated computational optics. No machine learning — pure wave propagation physics with intelligent enhancements.
1
Optical Configuration
Standard optical bench setup. Compatible with industrial cameras (8-bit depth) and coherent light sources.
2
Interference Pattern Capture
Record holographic fringes with standard industrial imaging sensors. No specialized hardware required.
3
API Submission
HTTP POST with holographic data. Platform accepts TIFF, PNG, NPY, RAW formats with optical parameters and enhancement parameters.
curl -X POST "https://api.apertureapi.com/reconstruct_pdhm?wavelength_um=0.532&pixel_pitch_um=3.45&z_min_mm=0.5&z_max_mm=3.0" \
-H "Authorization: Bearer $APERTURE_API_KEY" \
-F "file=@sample.tiff" \
-o reconstruction.zip
4
GPU-Accelerated Physics Processing
Angular spectrum propagation with autofocus and phase unwrapping. Platform includes: edge-aware bilateral filtering, SNR regime fusion (HIGH/MID/LOW), quality metrics. No AI/ML — pure computational optics.
5
Validated 3D Output with Quantitative Metrics
JSON response with quantitative phase and height maps, defect annotations, quality metrics, plus enhancement data (edge preservation ratios, SNR regime, fusion confidence). Every result carries its provenance and config commitment.
Paid validation pilots now open
A fixed-fee engagement with the acceptance threshold agreed before data is taken
Industry Verticals
Target applications: industrial quality assurance, semiconductor metrology, and precision optics. Edge-aware filtering is built for inspection tasks where sharp transitions matter.
Industrial Quality Assurance — Pilot Program
AdvancedCustomer Challenge: 2D inspection systems cannot detect critical 3D defects including BGA failures, solder volume errors, and microvia depth variations.
Our Solution:
- • 3D Solder Paste Inspection: Volume, area, height measurement before component placement (edge-aware detail preservation)
- • BGA Standoff Measurement: Collapsed solder ball detection (SNR fusion: intelligent path weighting)
- • Microvia Depth Verification: Non-contact laser-drilled hole measurement before plating (Zero-phase: <0.05 px drift)
Status
Paid validation pilots now open
A second, independent measurement point alongside your AOI system, on data you already have. Paid validation pilots now open.
Semiconductor Manufacturing
AdvancedCustomer Challenge: Wafer defects invisible to 2D inspection but critical to yield. Current metrology requires slow, expensive SEM or AFM.
Our Solution:
- • Wafer Defect Detection: Pit, scratch and contaminant identification on bare silicon (Edge-aware for sharp feature detection)
- • Etch Depth Metrology: Trench depth, feature height, sidewall angle measurement for process control (Zero-phase: no spatial drift)
- • MEMS Device Analysis: Real-time 3D profiling of cantilevers, membranes, micro-gears (Block overlap-add: MSE <1e-15)
Deployment Model
Enterprise integration
Off-axis architecture eliminates twin-image artifacts. SNR regime fusion optimizes dual-path processing. Non-destructive wafer inspection alternative to SEM.
Precision Optics
AdvancedCustomer Challenge: Surface quality determines optical performance. Contact profilometry risks damage; interferometry expensive and slow.
Our Solution:
- • Surface Roughness (Ra) Measurement: Non-contact topography of mirrors, lenses and optical flats (edge-aware filtering)
- • Micro-Lens Array QA: Radius of curvature, uniformity, height measurement of lenslet arrays (Zero-phase filtering)
- • Scratch/Dig Quantification: Precise depth/width measurement for MIL-SPEC or ISO compliance (Edge-aware detection)
Deployment Model
Production metrology
Full-color RGB holography reveals coating defects invisible to monochrome systems. Edge-aware enhancement preserves critical surface features.
Medical & Life Sciences
Customer Challenge: Traditional microscopy requires toxic stains that kill cells. Cannot observe living cells responding to drugs over time.
Our Solution:
- • Label-Free Live Cell Imaging: Phase shift measurement maps cell biomass and internal structure without staining
- • Drug Response Studies: Time-series observation of living cells reacting to stimuli
- • Medical Implant Surface QA: 3D roughness mapping for biointegration specifications
- • Clinical Diagnostics: Malaria detection, sickle cell, leukemia screening applications
Deployment Model
Research & clinical
Wide-field label-free imaging enables longitudinal studies of living biological systems
Platform Advantages
A deployment model built around committed artifacts: provenance, determinism, and an evidence report you can recompute.
🔒
Enterprise Security & Compliance
On-premises deployment available. Every result carries a provenance record, a content-addressed report ID and SHA-256 commitments to input and config, so a third party can verify what produced it.
FDA-suitable deterministic processing
💰
Leverage Existing Infrastructure
RESTful API integrates with existing MES and QA workflows. Works with standard industrial imaging sensors. Platform requires no hardware changes.
Integration over replacement
Software-Defined Architecture
AdvancedTraditional metrology requires complete system replacement when upgrading components. Our Advanced physics-based platform enables independent hardware evolution without vendor lock-in.
Legacy Approach
Integrated Hardware Systems
- • Single-vendor ecosystem dependency
- • Component upgrades require system replacement
- • Planned obsolescence every 3-5 years
- • Proprietary formats and limited integration
Our Approach
Physics-Based Software Platform
- • Deploy with any standard imaging sensor
- • Hardware evolution without software changes
- • Protected infrastructure investment
- • Open formats (TIFF, PNG, GLB, OBJ, STL)
- • Standard RESTful API integration
Platform Status
Recomputable Platform
- • Deterministic, recomputable results
- • A second, independent measurement point — beside your instrument, not instead of it
- • Edge-aware bilateral filtering
- • SNR regime fusion (HIGH/MID/LOW)
- • Benchmark regression gate enforced against a committed baseline
Where we are: paid validation pilots are now open. A pilot runs against your frames and your own reference measurement, with the acceptance threshold agreed before any data is taken — and all three committed. We have no production deployments to cite yet, and we will not imply otherwise.
Apply for Free Pilot ProgramVerified Facts
Every statement in this section is backed by a committed, independently checkable artifact. Where a figure has no committed artifact, we publish no figure.
Software-Defined Calibration — live in production
Digital-twin calibration is live on the production API, including the no-account /public/reconstruct endpoint. In the committed closed-loop proof, the optimizer recovered the true pixel pitch to within 0.0005% from a nominal that was mis-specified by 6%, and the digital twin replicated the measured field to within the optimizer's convergence tolerance. This is a closed-loop synthetic validation of optimizer convergence and endpoint plumbing — not a physical-target measurement, and we label it as such. Artifact: calibration/results/sdc_end_to_end_hex_nut.json.
Deploy provenance you can query
The live /health endpoint serves a full deploy-identity block, and a hash-chained deploy ledger verifies intact from genesis. Both are registered live checks (deploy-identity, ledger-verify) — ask, and the running system answers with a live run.
Data-retention governance, fail-closed
Retention policy is class-keyed with fail-closed owner routing (unclassified data routes to a named data-integrity owner, never to a default), write-once legal-hold events with declared release conditions, and provable clock-basis divergence. Verified by the registered retention-proof check.
Conformity decisions are data, not code
Quality-threshold decisions ship as content-addressed, supersession-chained records with an enforced separation between the business decision owner and the engineering computation. The freeze gate requires at least three independent evidence lineages — a passing verdict cannot be self-manufactured by the vendor.
Benchmark discipline
Every published performance number is stamped with the committed result artifact it came from; capabilities without a committed measurement are listed as not-yet-benchmarked and carry no number. A freshness check (bench-doc) fails the build if the published document drifts from the artifacts.
Regression budget enforced at every release
A quiet-window benchmark-regression gate runs against a committed baseline with a fixed 0.80 performance budget before changes are accepted. The gate result is a registered live check (bench-regression).
PrismView: Modular Hardware Reference Designs
Validated reference implementations for rapid deployment. Compatible with standard industrial imaging infrastructure. Software capabilities work with existing hardware.
Core Electronics Module
- Emitter Module: Tri-Color RGB Laser (638/520/450nm, 5mW) or high-power RGB LED for low-speckle applications
- CMOS Sensor: High-resolution monochrome sensor, pre-configured with 24-pin FFC connector
- Optics Module: 50/50 plate beamsplitter for off-axis configurations
- Aperture Module: Precision iris diaphragm for contrast control
- Calibration Standards: Reference objects for software-defined calibration
Reference design cost: $1,200-$1,800 depending on configuration
Integration Requirements
- Mechanical Assembly: Download STL/STEP reference files for custom chassis fabrication (PETG-CF or PLA recommended)
- System Integration: Mount electronics in chassis with plug-and-play FFC connectors (no soldering required)
- Manufacturing Partners: Pre-manufactured enclosures available from OEM partners
- Software Platform: All hardware configurations compatible with platform capabilities (no hardware changes required)
Complete reference implementation: $1,500-$2,500
Modular Component Architecture
Emitter and sensor modules are physically interchangeable across multiple configurations. Single module set enables deployment across diverse applications. Software capabilities work across all configurations.
Application A: Cell Biology (Advanced)
Deploy sensor and fiber-coupled laser in microscope chassis. Label-free live cell imaging at 130mm working distance. Edge-aware enhancement preserves cellular details.
Application B: Semiconductor QA (Advanced)
Reposition identical electronics into off-axis chassis. MEMS device inspection at 5mm object distance with artifact-free reconstruction. SNR regime fusion optimizes dual-path processing.
Inline Configurations
Direct optical path: Emitter → Object → Sensor.
- • Standard: 150mm emitter-to-object, 10-15mm object-to-sensor
- • Iris Aperture: Enables 5-10mm closer object placement for increased contrast
- • Microscopy: 10mm pinhole-to-sample, 130mm sample-to-sensor (wide-field)
Off-Axis Configurations
Beamsplitter creates separate reference path; eliminates twin-image artifact.
- • Compact: 200×150mm, 15-25° reference beam, 5-10mm object distance
- • Large-format: 600×400mm, 10-15° reference beam, 100mm object distance, 60×45mm field
Distance Parameter Impact
Object-to-sensor distance determines resolution-field tradeoff.
- • Close (5-15mm): High numerical aperture → micron-scale resolution, limited field
- • Far (100-130mm): Lower numerical aperture → extended field, coarser resolution
- • Software-Defined Calibration automatically corrects geometric variations
Platform Differentiation
Physics-based platform with a determinism contract — no AI training required.
Physics-Based — Not AI/ML
Computational optics based on wave propagation physics. No training datasets, no fine-tuning, no ML uncertainty. Deterministic results from day one. Platform maintains full determinism.
Recomputable Results
ValidatedQuantitative phase and height maps from data you already have. Edge-aware filtering and SNR-adaptive fusion.
Continuous Improvement
Platform improves through volume processing practice — the more data processed, the better the results. No ML training required for optimization. platform capabilities add customer-validated features with standard APIs.
Deployment Solutions
From free evaluation to paid validation pilot. Platform capabilities included at no additional cost.
🎉 Free Pilot Program
Full Platform
Risk-Free Evaluation
100% Free
- ✓ Process your holographic data at no cost
- ✓ Share anonymized results to help us improve (win-win)
- ✓ Full platform capabilities — no feature restrictions
- ✓ platform capabilities included (edge-aware filtering, SNR fusion)
- ✓ Comprehensive calibration profile for your hardware
- ✓ 60-90 day program with technical support
- ✓ No commitment to purchase required
Win-Win Model: You get free Advanced processing and validated 3D reconstructions. We get practice data to continuously improve our physics-based algorithms through volume processing.
Validation Pilot
Full Platform
Integration Partnership
Custom Pricing
- ✓ Volume-based processing tiers
- ✓ Production API with automated QA workflow integration
- ✓ Custom reconstruction parameter optimization
- ✓ platform capabilities: edge-aware, SNR regime fusion
- ✓ Multi-wavelength (RGB) and off-axis architecture support
- ✓ Priority processing
- ✓ Quarterly optimization reviews with optical engineering team
Scales from laboratory to multi-site production operations. Full platform included.
Contact Solutions TeamEnterprise Deployment
Full PlatformCustom Architecture
- ✓ Custom feature development and optimization
- ✓ On-premises or private cloud deployment options
- ✓ Dedicated engineering support resources
- ✓ platform capabilities with custom tuning
- ✓ Quarterly technical optimization sessions
- ✓ White-label licensing available
- ✓ Defect classification integration
For semiconductor fabs and tier-1 manufacturing. Full platform included.
Contact Enterprise TeamProduction API Access (Full Platform Included)
Development
$50/mo
1,000 reconstructions/month
- ✓ Complete API access
- ✓ Priority processing queue
- ✓ Technical documentation
- ✓ platform capabilities
Research labs, optical shops
Professional
$500/mo
20,000 reconstructions/month
- ✓ Priority processing queue
- ✓ Batch processing API
- ✓ Technical support
- ✓ platform capabilities
PCB fabrication, contract manufacturers
Enterprise API
$5,000/mo
Unlimited with dedicated support
- ✓ Negotiated availability terms
- ✓ Custom parameter optimization
- ✓ On-premises option
- ✓ Custom parameter optimization
Semiconductor fabs, tier-1 suppliers
Pay-As-You-Go
$0.03/scan
Zero monthly commitment
- ✓ Full API access
- ✓ No monthly minimum
- ✓ Production quality
- ✓ platform capabilities
Prototyping, low-volume production
All tiers include: platform capabilities • OpenAPI documentation • Evidence report per reconstruction • 14-day evaluation access
Physics-Based Processing Architecture
GPU-accelerated computational optics with deterministic quality assurance plus platform capabilities. No machine learning — pure wave propagation physics with intelligent edge preservation and SNR fusion.
Deterministic Reconstruction
ValidatedAngular spectrum propagation with automated focusing and phase unwrapping. Platform includes: edge-aware bilateral filtering, SNR regime fusion (HIGH/MID/LOW), quality metrics. Physics-based — not AI/ML.
Digital Twin Integration
Align digital representations with physical data. Version control for 3D models with collaborative access and audit trails. Comprehensive validation metrics included.
Enterprise Security
RESTful API with OpenAPI specification. Append-only result manifest, class-keyed retention policy with legal-hold support, on-premises deployment available.
Validated Physics-Based Processing
No AI/ML — deterministic computational optics. Every published performance number is stamped with the committed artifact it came from; capabilities without a committed measurement carry no number.
Angular Spectrum Propagation
Exact Helmholtz solution for numerical reconstruction. GPU-optimized for production processing. Physics-based — no machine learning. Platform enhances post-processing.
Automated Focal Detection
Automatic focal plane detection using variance, gradient, and Laplacian metrics. Standard 21-plane sweep with deterministic results. Platform includes zero-phase filtering (<0.05 px drift).
Phase Unwrapping
Quality-guided and Goldstein algorithms for validated 3D surface reconstruction from wrapped phase. Auditable for regulatory compliance. Platform maintains perfect reconstruction (MSE <1e-15).
Artifact Suppression
HIO, Gerchberg-Saxton, and Error Reduction methods eliminate twin-image artifacts. Validated reconstructions through deterministic physics. SNR regime fusion optimizes dual-path processing.
Edge-Aware Denoising
NewPlatform Capability: Advanced bilateral filtering preserves sharp transient features while eliminating sensor noise. Edge-aware by design. Exercised across diverse optical configurations in the test suite.
Universal Format Support
TIFF, PNG, NPY, RAW compatibility. Standard camera output accepted without proprietary format requirements. Open export formats (GLB, OBJ, STL, PLY). Validation metrics included in JSON response.
Start with Our Free Pilot Program
Full PlatformProcess your holographic data at no cost with full platform capabilities (edge-aware filtering, SNR regime fusion). Share anonymized results to help us improve through volume processing. No commitment required.
Win-Win: You get free Advanced processing and validated reconstructions. We improve through practice data. Physics-based platform — no AI training required. platform capabilities at no additional cost.
Technical Discussion
Discuss your metrology requirements with our optical engineering team. We'll walk through how the platform addresses your specific inspection challenges, and show you exactly what we can and cannot yet prove.