VisionCUT is a solution developed for automatic roll feed or sheet feed cutting processing, when this involves cutting a shape from a distorted media like: carpets, printed fabrics, airbags, labels, outdoor advertising, boat sails etc.

Usually, such material presents manufacturing distortions, due to the elasticity, dilatation and contraction, or the imprecision of printing, knitting or weaving process. If the cutter would follow the original precise cut-path, this would not follow the contour of the actual shape, and the result of cutting will be unacceptable.
The main goal of the application is to identify the position and distortion of the shape while the media is fed into the cutter, and to compensate for this by distorting the cut file within a given set of rules and restrictions.
The system contains a scanning module, installed on the feeding area of the cutter. The scanning module can be installed on virtually any type of conveyor cutter because the space required for scanning is quite small, only a 5 centimeters linear frame is enough to capture the image of the material while it is feeding.
While the material is fed on the conveyor, that scanner acquires the surface image and transmits it to the processing software. The software identifies the actual position and distortion of the shapes on the material, by matching markers on the surface with predefined markers from the cad file. The markers can be crosses, continuous lines, points etc., depending on the field of application. The software then compensates the shifting in position and distortions by altering the original cut file to match perfectly the new contour. This alteration is made under specific rules and restrictions defined by the user.
The altered cut file is sent to the cutter in real time, and the result is a perfect match between the cut line and the actual shape in the material.
The same system can be used to slice a material along its length on a specific split line, when this line is variable, or on other specific application that requires real-time media inspection.