What can contribute an artificial vision system?
Endow to quality controls with high accuracy and reproducibility. They are systems able to recognize, through various sensors, all variables that govern a process. All of it without the need of human intervention.
Being able to detect and differentiate different elements using codes OCR or 2D codes in assembly lines. Even for simple elements is able to detect and classify according to shapes through recognition algorithms.
The main advantage of these systems. Increase the efficiency of production lines and production processes providing 'eyes’ to workstations, through sensors and cameras. It is a way to add intelligence to the production process, being able to react to changes, foreseen or unforeseen, and auto-regulate.
Types of artificial vision
How does an artificial vision system behaves?
1. Image Acquisition
At first the system sensors (visual, infrared, magnetic field,…) acquire information from the outside, what we call 'image'.
2. Database reading
It is the memory system. It can be libraries, images, and patterns inter alia.. It finds the proper pattern to compare with the image.
Suitable algorithms for each application are responsible to process image and pattern.
4. Create outputs or system reaction
According to the result of comparing it will create the system outputs that have been previously defined. Or in the case of more advanced systems those tha the system has learned.