In the context of manufacturing, machine vision is a branch of engineering that uses computer vision. Machine vision analyse images to extract data for controlling a process or activity. Machine visions are applied for the quality assurance, sorting, material handling, robot guidance and calibration.
For any vision system, 2 important specifications are sensitivity and resolution. Sensitivity here refers to the ability to detect or differentiate small changes like brightness. As for resolution, it’s the ability to differentiate objects. Usually, sensitivity and resolution are interdependent. This means if sensitivity is high, resolution will be lower and vice-versa.
One of the areas where machine vision system is used is the 3D Model Construction. This is extensively used in entertainment (animation and effects) industry to create digital 3D models. Using scanned 3D models is a lot easier and time saving compared to creating 3D models from scratch.
3D scanner works by scanning and analyzing a real object and reconstruct it as a 3D model in digital mode. Scanning and analyzing is done when the scanner projects wave to the object(s) and collects the reflection of the wave. This wave can be light, ultrasound, laser or x-ray.
An example of a 3D scanner is the time-of-flight laser scanner, a non-contact active scanner type. The main part of this scanner is the laser rangefinder. Just like the name itself, rangefinder, it finds the range between the scanner and object by deducing time for the laser to reflect from the object to the scanner. The precision to measure time is very important for this device because it signifies accuracy.
A laser rangefinder detects an object one point at a time, which means to detect an object; direction view of the scanner must change to scan different point. To change direction view, we either rotate the rangefinder or use a rotating mirror system. Rotating mirror system is the better choice because they’re lighter, making them faster to rotate at a better accuracy.
Another example where machine vision system is used, is for quality assurance. It is the systematic monitoring and evaluation of the various aspects of a project, service or facility to maximize the probability that minimum standards of quality are being attained by the production process. Quality assurance cannot absolutely guarantee the production of quality products but it can reduce the probability of damage. Quality assurance will make sure that the quality of the product produced at least at the minimum requirement for the production.
The system involve camera for motion capture and process the image captured into digital form of image. The system works by capturing the image of the output product and compare the image with the set point image. If the image captured did not satisfy the requirement, the system will trigger the product as faulty and thus the product will be sent back.
For example, label inspection on wine bottles. If the figure/label on the bottle is not the same in database, then the bottle will be sidelined as reject.
There are other applications of machine vision system like in the security (e.g retina recognition, thumb print recognition), medical (bacteria detection) and safety (currency detector)sector.
No comments:
Post a Comment