Predictive Quality Control
with RTI Vision Sensor solino®
Digitising the Visual Appearance of Surfaces
solino® transforms how quality control is performed. Instead of selective sample testing, it enables continuous and automated surface monitoring across a large field of view with Reflectance Transformation Imaging (RTI). All anomalies are detected with mathematical precision. By combining BRDF-based optical detection with intelligent algorithms, this sensor captures the true visual appearance of materials – revealing anomalies beyond human perception. It works reliably under normal factory lighting, remains robust against stray light and transforms visual impressions into measurable perfection.
solino® sets a new benchmark for automated predictive surface quality control in precision manufacturing.
With solino®, for the first time ever, you can:
- digitise the visual impression of any surface
- assign a unique digital fingerprint to every sample
- benefit from a Computational Imaging Sensor that combines hardware and software
- integrate a Plug & Play sensor, ready for use in minutes
- evaluate statistically and reproducibly
- eliminate reflections even on shiny metals
- predict product failure through AI correlation
solino® Technology
solino® is based on Bidirectional Reflection Distribution Function (BRDF) measurement – the physics behind visual appearance.
It uses 64 LEDs in a precisely calibrated dome to illuminate the surface from all directions. The reflected light is analysed to derive:
- colour, waviness and roughness
- reflectivity and inclination
- surface stress and microstructure anomalies
This light-scattering pattern becomes a unique optical signature – the ideal basis for automated, repeatable and objective inspection.
How solino® works
- Image Acquisition
Photometric stereo setup captures multiple views of the same surface using 64 LEDs with known position, incidence and intensity. - Compute specific images with BRDF
Out of the 64 images set, solino calculates specific images – revealing new visual data and removing irrelevant reflections.
Each image isolates particular characteristics (e.g. gloss, texture, structure). - BRDF Data Extraction
Consistent capture and calibration provide full BRDF information, creating a reliable data foundation for further AI analysis. - Anomaly Visualization & Analysis
Clean, repeatable and interpretable CPIs for generating anomaly heatmaps in AI image processing software.
Sample Gallery
Any further questions about solino® in your specific application.
We’re here to discuss your application or arrange a live demo with your samples.
FAQ
solino® can communicate using standard industrial interfaces (I/O, Ethernet IP, OPC UA, etc.).
Automation is easily implemented: solino® can be automatically triggered to make inline inspection, enabling evaluating results and logging data in your MES or QMS system.
- 100 % Quality Control = 100 % Reliability
- lower equipment and labor costs
- full 24/7 automation capability
- new, objective, digital quality standards
The solino® image acquisition is based on the perfect calibrated knowledge of the angles of the light source and the optical axis of the camera. The primary goal is to be able to detect the minimum scattering pulse for a wide variety of objects. This is the only way to detect all defects, always and on all objects. However, this requires a very close-meshed arrangement of LEDs - 64 white light LEDs in a dome arrangement.
solino® processes the image information of each lighting scenario and processes it with information about:
- the position and collimation
- the distance and colour
- the type of source
- the stray light component
It´s based on a industrial camera and a multiple LEDs dome.
'Photometric Stereo' is the acquisition of images taken from different angles of illumination.
The sensor does not differ in image acquisition, but it does differ significantly in image processing. Instead of working with the individual image information, the scattering value per object point is calculated two-dimensionally and revealing new information in the images - and thus anomalies can be evaluated and recognized using an AI processing software.
Although solino® images often look the same, the image information they contain is:
- created by evaluating scattering information on the surface and not by images alone
- repeatable and comprehensive, as the 64 LEDs illumination reveals all possible defects on all possible surfaces
- object-specific adaptable, with which anomalies (defects, scratches, etc.) can be specifically elaborated
Perfect for classifications and as a data source for AI software packages.
solino® can be seamlessly integrated into existing production or inspection workflows. We provide SDKs, APIs and integration support for inline inspection systems, conveyor setups, or robotic applications.
solino digitized the optical footprint (BRDF) of a surface that then can be classified and trained with a AI image processing software - ready automation.
The scattering behavior of a surface is based on its structure. It is digitized with solino using the BRDF - the ideal digital footprint of a surface. In the solino® data set, all relevant object information is stored:
- geometric shapes and positional deviations
- colour, waviness and roughness
- anomalies
- etc.
The visual appearance of objects is determined by the way they reflect light.
The colour of objects is determined by the parts of the light spectrum that are reflected or transmitted without being absorbed.
We can also say that the appearance characteristics of an object are based on the directional distribution of reflected light (BRDF) - the specific image result from solino.
The appearance could be described by attributes such as dull vs. high gloss.
Dull
- does not reflect light and is less glossy
- softer images and colors with a wider tonal range
- soft and supple to the touch
High gloss
- highly reflective for radiant shine
- more vivid, saturated colors that highlight details
- glossy, wet look that provides a striking appearance
solino® is based on specially developed algorithms that analyze the scatteringe of incident light on surfaces.
These in turn are displayed as individual images that show the change as a contrast or colour revealing ununiformity, e.g., anomalies or defect.
Thanks to solino®, the images acquired are cleaned, so to speak - slight object anomalies are highlighted and can be displayed as clearly dominant image information.
We consider the appearance of the surface as a combination of light and three other production parameters:
- material
- shape
- tint
The visual image information collected is the perceived colour as a result of the interaction between light and surface.
The LED illumination is a known parameter, as is the camera sensor.
solino® defines relative quality rules that depend on the hardware but are based on human perception. The goal is to obtain an automated inspection system that digitizes visual perception.
solino® correlates the light positions
Based on calibrated illumination angles and intensity profiles, the solino® algorithm shows all surface anomalies.
The decisive factor is not the type, intensity or homogeneity of the light source used. The assignment of the image acquisition to the light identity is the most important part. In order to be able to achieve optimal results, special tools for system calibration have been developed accordingly.
Computation of particular Images (CPI) is a software solution for calculating a (specific) image from a whole image set. Consistency and an exact system calibration are the top priority for the image input.
Due to this acquisition consistency, BRDF information results are decomposed into sub-layers with an image as a sub-layer.
Each (particular) image is the result of several processing steps - processing from the distribution of pixel values.
To improve quality control processes, CPI offers a surface anomaly inspection image gallery as a first step up to automated inspection driven by humans and supported by AI.
Representation of product features as image information
When identifying and determining surface defects, it is essential to be able to recognize the defect in the image in the first place or to be able to find it again and again, regardless of the ambient conditions. The same applies to similar anomalies, which must also be detected and displayed accordingly.
solino® was developed for this purpose. Image information that is not relevant is deliberately suppressed and anomalies are highlighted.
Image cleaning of unimportant factors is performed, resulting in repeatable and reliable image analysis.
solinoStudio user interface
Yes, the solinoStudio supplied with our solino® allows a live image display as well as a preview of four processed CPI anomaly images.
A total of more than 200 algorithms for analysis and evaluation are available for the 64 individual exposures, according to which the result images are displayed. The selection is made using the drop-down menu (see image on the right).
The solinoViewer offers:
- an intuitive user interface to identify the product defects or anomalies
- a slider image manager with one widget per image
- the solino Image Editor for batch processing of stereo-photometric images
- a selective list of image processing processes to be output
- the selection of the color channel to be processed
Depending on the computer (power, RAM, processor, etc.) it could be less than 5 sec.
Our solinos sensors are made for samples beween 10x10 to 50x50 millimeters. Any kind of surface can be tested and evaluated.
The main limit is the samples size. Its limited by the size of the illumination or dome. For samples outside the standard range, please contact sales@opto-gmbh.com.