Matrox Imaging made a major update to Matrox Design Assistant X, the flowchart-based vision software. With a field-proven record, this integrated development environment (IDE) provides an intuitive way to develop vision applications without writing traditional program code.
The latest release delivers a range of additions and enhancements, including interfaces to additional 3D sensors; expanded image classification using deep learning; image capture from CoaXPress cameras; and expanded support for working with programmable logic controllers (PLCs).
3D sensor interfacing, processing, and analysis
Matrox Design Assistant X now readily interfaces with the new Matrox AltiZ 3D profile sensors, as well as 3D sensors from third-party manufacturers. The software works seamlessly with these devices to process and analyze depth maps. Matrox Design Assistant X also has the capabilities to analyze 3D profiles obtained directly from a sensor or extracted from a point cloud or depth map.
Coarse segmentation using deep learning
The classification steps offered in Matrox Design Assistant X make use of machine learning to categorize image content. The new CNNClassMap step uses deep learning technology to perform coarse segmentation, generating maps indicating the pre-established class and score for all image neighborhoods. This new step, in addition to the other classification tools, provides users with key means of detecting defects in highly textured, naturally varying, and acceptably deformed materials.
Users of Matrox Design Assistant X can opt to train the convolutional neural network (CNN) on their own, making use of the MIL CoPilot interactive environment available separately as part of Matrox Imaging Library (MIL) X software. MIL CoPilot lets users readily develop the required training dataset, including the labeling of images, augmenting the dataset with synthesized images, and monitoring and analyzing the training process. Alternately, users can commission Matrox Imaging’s team of experts to conduct the training on their behalf.
Image acquisition from CoaXPress cameras
Matrox Design Assistant X already captured images from any GigE Vision or USB3 Vision compliant camera, and the latest version now lets users capture images directly from one or more CoaXPress cameras using a Matrox Rapixo CXP frame grabber.
Direct connection to PLCs
Version 2006 of Matrox Design Assistant X now offers CC-Link IE Field Basic, along with EtherNet/IP, Modbus over TCP/IP, and PROFINET industrial communication. CC-Link IE Field Basic communication enables direct interaction with programmable logic/automation controllers from Mitsubishi Electric®. Moreover, the flowchart-based software now also features a PLC interface emulator so that users can test communication with a PLC without needing one at hand, thereby simplifying the application development and maintenance process.
Software update for Matrox Imaging smart cameras and vision controllers
This latest software update extends to the Matrox Iris GTR smart cameras and Matrox 4Sight EV6 vision controllers bundled with Matrox Design Assistant X, expanding the features, functionality, and connectivity of these capable machine vision platforms.
“This update affords users of Matrox Design Assistant X great new additions and enhancements,” says Fabio Perelli, product manager, Matrox Imaging. “By extending the support of 3D cameras and PLCs, plus integration of even more deep learning technology, Matrox Design Assistant X continues to deliver a flowchart-based environment that streamlines machine vision application development.”