Description
Lanit-Tercom Italia develops computer vision and artificial vision systems capable of transforming images into streams of information understandable for applications, recognizing and classifying objects and perceiving the spaces around observation points. The experience gained in numerous projects with clients worldwide has led to computer vision solutions applicable in various contexts: from industrial quality control to road safety, from construction site management to urban monitoring.
Computer vision systems combine convolutional neural networks, semantic segmentation models, and multispectral processing techniques to detect defects, objects, living beings, and their behaviors in real-time, even in extreme environmental conditions. Each computer vision project stems from an approach that combines mathematical rigor, technological independence, and customization, ensuring certifiable results in terms of accuracy and processing speed.
Applications for Industry and Quality Control
Specifically in industrial quality control, our computer vision systems (such as ProdScan) automate the identification of defects in manufactured products.
Leveraging advanced visual analysis based on computer vision—which includes high-definition, polarized, or infrared cameras and machine learning algorithms—these solutions verify full compliance of surfaces and geometries. This approach allows for the detection of minimal imperfections, often invisible to the naked eye, ensuring high quality standards, reducing waste, and integrating with factory management systems (MES/ERP).
Applications for Mobility and Urban Management
For road safety and the automotive sector, we develop computer vision solutions (such as ADAS Vision) for real-time analysis of the road scene. These systems are designed to operate effectively even in low-visibility conditions such as fog, rain, or night driving.
Using advanced computer vision techniques and semantic segmentation, the software detects, tracks, and classifies people, vehicles, horizontal and vertical signage, and obstacles. The same rigor applies to urban monitoring (with platforms like GeoAI), where computer vision and deep learning solutions analyze georeferenced images to automatically map and classify infrastructure such as streetlights, trees, or waste.
Applications for Specialized Sectors (Agriculture and Industry)
Our expertise in computer vision extends to specific fields such as precision agriculture and industrial safety. In agriculture (with VineSense), computer vision is integrated onto machinery to monitor vineyards, using neural networks (such as YOLOv4 and Unet) to estimate the volume of grape clusters in real-time.
In the industrial field (with SafeOps), computer vision systems monitor compliance with safety regulations, automatically detecting the correct use of PPE or the presence of personnel in hazardous areas, dynamically adapting rules based on signals from corporate systems (SCADA).
