The name of the base system (platform): | Vmx Dequs: ISE |
Developers: | VideoMatrix (Videomatrix) |
Date of the premiere of the system: | 2020/04 |
Branches: | Metallurgical industry |
Technology: | Video surveillance system, Video analytics systems |
Content |
The main articles are:
- Computer vision (machine vision)
- Video analytics (terms, applications, technologies)
- Neural networks (neural networks)
2022: Automation of quality analysis of military steel sheets using VideoMatrix technology
On July 29, 2022, VideoMatrix announced that it had patented the technology of flaw detection of reflective surfaces with machine vision. Using the patented technology based on the Vmx Dequs software, Rostec Group of Companies automated analysis of the quality of military steel sheets at the ARMIL software and hardware complex. Read more here.
2020: Development of ARMIL for automated analysis of steel sheets for defects
In April 2020, it became known that the company "VideoMatrix" , together with specialists from RT-Tekhpriemka (state corporation) Rostec , developed ARMIL (Automated Workplace for Inspection of Steel Sheets), a software and hardware complex for steel inspection (in particular, VNS-9Sh steel, which is used in the creation of combat and civilian helicopters). The complex combines neural network technologies and. computer vision
Finished steel sheets are placed on the ARMIL platform and undergo multispectral analytical processing based on VideoMatrix's own Vmx Dequs software. The system automatically detects more than 20 classes of defects, including scratches and microcracks measuring from 0.3 mm with an accuracy of 97%.
Applied technologies:
- The technology is based on interference-resistant algorithms (statistical analysis, graph theory, spatial filtering, its own algorithm for extracting objects).
- Vmx Dequs software is based on technology artificial intelligence and mathematical technologies. machine vision The resulting video stream is analyzed. server video analytics The sensor components are cameras of visible spectrum, MC, laser. scanners
- The system compares the steel surface with the reference and, based on the specified characteristics, detects product defects and classifies them. For this, the algorithmic detection method is used. For a more accurate result, the operation of detecting and classifying neural networks is superimposed.[1]