AI Gateway and Inspection plug-in Pleora Technologies and Neurocle have developed the AI Gateway and Inspection plug-in for simplifying the deployment of deep learning-based classification, segmentation, and object detection for visual inspection. Te plug-in allows end-users and

integrators to deploy machine learning visual inspection capabilities without any additional programming. Images and data are uploaded to Neurocle’s Neuro-T, deep learning vision software that supports an auto deep learning technology with predefined parameters optimised for the Nvidia GPU in Pleora’s AI Gateway. AI models are transferred and deployed on the AI Gateway for production environments. Te AI Gateway handles image acquisition

from any vision standard-compliant image source and sends out the processed data over GigE Vision to inspection and analysis platforms. In a quality inspection application, for

Beta version 5.0 of Adaptive Vision Studio Adaptive Vision has introduced beta version 5.0 of its software product portfolio, including Adaptive Vision Studio. Te release has new features of HMI events and parallel tasks, which make it possible to create larger and more interactive applications. Moreover, a new formulas engine gives easier analysis of inspection results. Te application now also features a streamlined program edition mode for simple applications, with pre-defined program sections, minimal program view and simplified result analysis in the results window. Te release also brings


One millimetre optical filters Schneider-Kreuznach has launched 1mm optical filters, designed to be placed between lens and sensor. Equipped with a Schneider- Kreuznach coating with high transmission and steep slopes, the filters are an ideal choice for challenging machine vision applications. Te 1mm thin filters are provided with an ultra- low reflection coating, important for traffic surveillance, for example. Te wavelength tolerance of ±1 per cent makes them a

preferred choice for metrology systems. Optimised for common LEDs used in

machine vision, the filters are ideal for automated inspection systems, such as those used in food and beverage control or logistics. For example, the blue BP 465-70 HT bandpass is often used in 3D measurement applications. For night and day vision cameras, the company offers the VIS-85 type filter, transmitting in the visible from 430nm to 680nm and 855nm, ±15nm.


example, the AI Gateway intercepts the camera image feed and applies the selected plug-in skills. Te gateway then sends the AI processed data to the inspection application, which receives the video as if it were still connected directly to the camera. Similarly, the AI Gateway can process imaging data with loaded plug-in skills in parallel to traditional processing tools. If a defect is detected, processed video from the

advances in deep learning technology, introducing full support for Adaptive Vision’s Weaver inference engine by all anomaly detection and instance segmentation tools, as well as improved optimisation for Nvidia RTX series GPUs and CPU back- end of Weaver. Finally, version 5.0 comes with an optimised reader for data matrix and QR codes. Adaptive Vision also

introduced version 5.0 of Adaptive Vision Library, now based on modern versions of C++. Previous support for Microsoft Visual Studio 2013 has been dropped, replaced by support for versions 2015 and later.

AI Gateway can confirm or reject results as a secondary inspection. Pleora’s AI Gateway provides additional plug-

in AI skills for hyperspectral imaging, with the processing flexibility of an Nvidia GPU to deploy open source or custom algorithms developed in popular frameworks like TensorFlow and OpenCV.

Halcon 20.05 MVTec Software has launched a Progress release of its standard software Halcon. Te barcode reader in Halcon 20.05 has been improved by an advanced decoding algorithm, which increases the decoding rate when reading codes with very thin bars. Tanks to this, it is now possible to read codes with bars smaller than one pixel. Training for all deep

learning technologies can now be performed on a CPU with the new software. Te generic box finder has been improved in terms of robustness, performance, speed, and usability. In addition, the Grad- CAM-based heatmap – gradient-

weighted class activation mapping – can now be calculated on a CPU without significant speed drops. Tis means that users are able to analyse directly which parts of an image influence the deep learning network’s classification decision. Moreover, anomaly detection

has been improved. Training the network is now up to 10 times faster, and inference speed has also been improved. Trained networks now also requires less memory and disk space, which makes the feature more viable for use on embedded devices. Halcon’s surface-based 3D matching has also been improved.

Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42