Detect objects in spectral images and classify objects by content

Major release of perClass Mira brings many enhancements.

Visualize plant health on plant leaves using spectral indices in spectral images

The easiest portable lab-scanning solution for spectral imaging

Visualize plant health on plant leaves using spectral indices in spectral images

perClass joins Headwall Photonics

Visualize plant health on plant leaves using spectral indices in spectral images

Visualize plant health in minutes

Interactively created classification solution for hyperspectral data

Quickly create robust sorting solutions

Estimate tomato brix using hyperspectral imaging

Quantify sugar content in tomatoes

perClass Mira is the easiest interface for spectral imaging with real-time deployment

Spectral cameras enable new types of applications sensing material composition. perClass Mira empowers you to interpret spectral images in minutes and create robust and accurate solutions.

Key benefits

Upcoming events

Partners

Mastor KoreaSenop hyperspectral cameras
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
<b>perClass Mira</b> allows anyone to build interpretation solutions for spectral images in an easy-to-use GUI.<br/><b>For whom?</b><ul><li>Industrial practitioners and domain experts can build solutions detecting user-defined materials/defects in spectral images without programming or machine learning know-how</li><li>Researchers and R&D specialists can quickly solve well-defined problems like object segementation, extration of leaves on plants or definition of area of interest by clicking only saving time</li></ul> <b>perClass Mira Runtime</b> helps to quickly deploy models build in perClass Mira to production i.e. on a live stream of spectral data. It provides functionality specific to line-scan spectral imaging systems. Namely:<ul><li>processing model for individual spectral frames in a line-scan system (BIL layout)</li><li>multiple computational backends including GPU (NVidia and OpenCL) and multi-core CPU</li><li>correction of raw data by user-defined dark and white references</li><li>extration of connected objects and their coordinates for actiator</li></ul>perClass Mira Runtime contains also the perClass Runtime library. <b>perClas Mira Dev</b> bundles the GUI with software development kit (SDK) and perClass Mira Runtime enabling design and deployment for interpretation solutions to custom applications.<br/><b>For whom?</b>For practitioers who want to design classification solutions by clicking and deploy them in a production (sorting system, plant-phenotyping solutions, recycling applications etc.) <b>perClass Toolbox</b> is a Matlab toolbox with a rich set of practically-oriented tools for building industrial machine learning solutions. It does not require any other Matlab toolbox. <br/><b>For whom?</b> For a researcher who wants to have a full control on machine learning system design (and wants to publish their algorithm designs) Basic understanding on machine learning design methodology is needed (perClass provides regular training courses) <b>perClass Runtime</b> provides general environment for deployment (execution) of trained machine learning models on new data. It can be easily called from any application that can call a DLL library. <b>perClass Pro</b> is a bundle enabling R&D specialists to design custom machine learning solutions in Matlab and deploym them to custom applications without Matlab. <br/><b>For whom?</b> For researchers who want to put their custom algorithm designs into a live production system. <b>perClass Mira Pro</b> combines both the GUI and full-control Matlab-based tools with a development SDK and runtime library.<br/><b>For whom?</b> For researcherers who want to save significant amount of time with a GUI allowing them to solve parts of their solutions with automatic machine learning and wish to have full control on that one specific task where their expertise is needed. They also want to deploy the entire solution to an industrial process on new data.