Thorium Model Vision™ (TMV™) is a powerful machine vision software system for advanced object recognition and automated inspection.
TMV can process video and images from any standard video or image source, particularly live video feeds connected to a computer.
While TMV is built with a simple user interface, it also provides a wide range of customization and specific settings for advanced users and system designers.
We are now introducing a new 3D matching system using a single 2D camera.
This system allows real-time 3D computer vision without the limitations of 3D cameras or range-finding systems.
Features of our new 3D system include:
- Object detection and pose determination
- Accurate distance estimation
- Real time tracking
TMV uses a complex set of algorithms to extract information about an image. It combines low level processes, such as edge detection, with more sophisticated analysis methods. Reducing the image to regions of uniform color and intensity is just one example.
Creating a model of an object for the program to detect is as simple as placing a sample object in view of the camera and saving it for future use.
TMV uses these models to identify objects, as well as to extract information about an object's position, orientation, and size.
By using multiple models, the program can find different objects in the field of view or multiple instances of the same object.
TMV has a variety of applications.
TMV can be used to inspect for defects by creating a model and designating one or more parts to detect if they are present, absent, or defective.
In addition, areas of a model can be designated to look for inclusions such as stains, chips, or other unwanted manufacturing defects.
The user can define specific tolerances for a part's size and alignment and the program can detect when tolerances are exceeded.
TMV was created for flexibility to meet your particular needs. Contact Thorium Vision Corporation today to see how we can apply the power of TMV to your computer vision application.