Different types of vision systems

Selecting the right vision system is critical to meet your specific vision applications. Generally speaking, the different types of vision systems include:

1D vision systems

1D vision analyzes a digital signal one line at a time, rather than viewing the image as a whole, such as evaluating the variation between the most recent group of ten lines analyzed and a previous group. This technique typically detects and classifies defects in continuously manufactured materials such as paper, metal, plastic, and other nonwoven sheet or roll goods.

2D vision systems

The most common inspection cameras perform area scans that involve taking 2D snapshots of different resolutions. Another type of 2D machine vision, linear scanning, builds a 2D image line by line.

Comparison between area and linear scanning

In certain applications, linear bzo scanning systems have specific sales over area scanning systems. For example, inspecting round or cylindrical parts may require multiple area scan cameras to cover the entire surface. However, rotating the part in front of a single linear scan camera captures the entire surface when unwrapping the image. Line scan systems are best suited to small spaces for those cases where the camera must be positioned between rollers on a conveyor belt to see the bottom of a part. Line scan systems can also provide higher resolution than traditional cameras. Since linear scanning systems require moving parts to create the image, they are often suitable for continuously moving products.

3D vision systems

3D machine vision systems typically encompass multiple cameras, or one or more laser displacement sensors.

Multi-camera 3D vision in robotic guidance applications provides the robot with information regarding part orientation. These systems feature multiple cameras mounted at different locations and triangulated over the position of a target in 3D space.

In contrast, 3D displacement laser sensor applications typically include surface inspection and volume measurement, generating 3D results with just one camera. An elevation map is generated from the location of the offset of the reflected lasers on an object. It is necessary to move the object or camera to scan the entire product, similar to linear scanning. With a calibrated offset laser, displacement sensors can measure parameters such as surface height and flatness with an accuracy within 20 μm. Figure 15 shows a 3D laser displacement sensor inspecting the surface of a brake pad for defects.